1. DL manuals
  2. IBM
  3. Software
  4. High Performance Storage System HPSS
  5. Management Manual

IBM High Performance Storage System HPSS Management Manual

Other manuals for High Performance Storage System HPSS: Installation Manual, Installation Manual
Manual is about: High Performance Storage System Release 4.5

Summary of High Performance Storage System HPSS

  • Page 1

    Hpss management guide high performance storage system release 7.3 november 2009 (revision 1.0) hpss management guide november 2009 release 7.3 (revision 1.0) 1.

  • Page 2

    © copyright (c) 1992, 2009 international business machines corporation, the regents of the university of california, los alamos national security, llc, lawrence livermore national security, llc, sandia corporation, and ut-battelle. All rights reserved. Portions of this work were produced by lawrence...

  • Page 3

    Table of contents chapter 1. Hpss 7.1 configuration overview.....................................................................................15 1.1. Introduction...................................................................................................................................15 1...

  • Page 4

    3.3.3.1. Login.Conf........................................................................................................................................41 3.3.3.2. Krb5.Conf (for use with kerberos authentication only)................................................................41 3.3.4. Ssm he...

  • Page 5

    4.2.4. Modifying a storage subsystem..............................................................................................................81 4.2.5. Deleting a storage subsystem..................................................................................................................81...

  • Page 6

    5.1.1.2. Migration/purge server information window..............................................................................133 5.1.1.3. Mover information window.........................................................................................................133 5.1.1.1. Physical volume l...

  • Page 7

    6.2.4. Deleting a storage hierarchy definition...............................................................................................173 6.3. Classes of service........................................................................................................................174 6.3.1. C...

  • Page 8

    8.1. Adding storage space..................................................................................................................223 8.1.1. Importing volumes into hpss .............................................................................................................223 8.1.1.1. ...

  • Page 9

    9.1. Logging overview.......................................................................................................................294 9.2. Log policies.................................................................................................................................295 9.2.1....

  • Page 10

    13.1. Managing hpss users...............................................................................................................325 13.1.1. Adding hpss users.............................................................................................................................325 13.1.1...

  • Page 11

    15.1.2. Overview of the db2 backup process................................................................................................357 15.1.2.1. Configuring db2 for online backup..........................................................................................358 15.1.3. Overview of t...

  • Page 12

    List of tables table 1. Ssm general options..............................................................................................................39 table 2. Hpssgui specific options.....................................................................................................40 table 3...

  • Page 13

    Preface who should read this book the hpss management guide is intended as a resource for hpss administrators. For those performing the initial configuration for a new hpss system, chapter 1 provides a configuration roadmap. For both new systems and those upgraded from a previous release, chapter 1 ...

  • Page 14

    Hpss management guide november 2009 release 7.3 (revision 1.0) 14.

  • Page 15

    Chapter 1. Hpss 7.1 configuration overview 1.1. Introduction this chapter defines the high-level steps necessary to configure, start, and verify correct operation of a new 7.1 hpss system, whether that system is created from scratch or created by conversion from a 6.2 hpss system. To create or modif...

  • Page 16

    Listed. Each step is required unless otherwise indicated. Each step is discussed in more detail in the referenced section. 1. Configure storage subsystems (section 4.2.2: creating a new storage subsystem on page 76) subsystems can be configured only partially at this time. The gatekeeper , default c...

  • Page 17

    B. Create storage resources (section 8.1.2: creating storage resources on page 234) 4. Create additional hpss users (section 13.1.1: adding hpss users on page 325) 5. Create filesets and junctions (section 10.1: filesets & junctions list on page 308 and section 10.5: creating a junction on page 315)...

  • Page 18

    • verify that a core server and migration purge server have been configured for each storage subsystem. • verify that each storage subsystem is accessible by using lsjunctions and ensuring that there is at least one junction to the root fileset of each subsystem. (the root fileset for a given subsys...

  • Page 19

    Chosen if specified by their cos id . • verify that classes of service with multiple copies have the retry stage failures from secondary copy flag enabled. File families, filesets, and junctions • verify that file families and filesets are created according to the site’s requirements. • verify that ...

  • Page 20

    • monitor free space from the top level storage class in each hierarchy to verify that the migration and purge policy are maintaining adequate free space. 1.6.3. Performance checklist measure data transfer rates in each cos for: • client writes to disk • migration from disk to tape • staging from ta...

  • Page 21

    Chapter 2. Security and system access 2.1. Security services as of release 6.2, hpss no longer uses dce security services. The new approach to security divides services into two apis, known as mechanisms, each of which has multiple implementations. Configuration files control which implementation of...

  • Page 22

    This can be "unix" or "ldap". · - a string used by the authorization mechanism to locate the security data for this realm. This should be "unix" for unix authorization, and for ldap it should be an ldap url used to locate the entry for the security realm in an ldap directory. 2.1.2. Security mechani...

  • Page 23

    2.1.2.3. Ldap ldap authorization is not supported by ibm service agreements. The following information is provided for sites planning to use ldap authorization with hpss 7.1 as a site supported feature. An option for the authorization mechanism is to store hpss security information in an ldap direct...

  • Page 24

    To create a new group, use the following command at the hpss_ldap_admin prompt: group create -gid -name [-uuid ] if no uuid is supplied, one will be generated. • deleting a group to delete a group, use the following command at the hpss_ldap_admin prompt: group delete [-gid ] [-name ] [-uuid ] you ma...

  • Page 25

    Obtained from the foreign site's administrator. An example would be: "ldap://theirldapserver.Foreign.Com/cn=foreignrealm.Foreign.Com" • deleting a trusted foreign realm to delete an entry for a trusted foreign realm, use the following hpss_ldap_admin command: trealm delete [-id ] [-name ] any of the...

  • Page 26

    Rw---dt user ${hpss_principal_pvr} rw-c-dt user ${hpss_principal_ssm} ------t any_other pvr: rw---dt user ${hpss_principal_pvl} rw-c--t user ${hpss_principal_ssm} ------t any_other ssm: rwxcidt user ${hpss_principal_adm_user} ------t any_other all other types: rw-c-dt user ${hpss_principal_ssm} ----...

  • Page 27

    2.4.1. Configuring/updating a location policy the location policy can be created and updated using the location policy window. If the location policy does not exist, the fields will be displayed with default values for a new policy. Otherwise, the configured policy will be displayed. Once a location...

  • Page 28

    Maximum request threads. The maximum number of concurrent client requests allowed. Advice - if the location server is reporting heavy loads, increase this number. If this number is above 300, consider replicating the location server on a different machine. Note if this value is changed, the general ...

  • Page 29

    1. Add the hpss_restricted_user_file environment variable to /var/hpss/etc/env.Conf. Set the value of this variable to the name of the file that will contain the list of restricted users. For example: hpss_restricted_user_file=/var/hpss/etc/restricted_users 2. Edit the file and add the name of the u...

  • Page 30

    Field descriptions restricted users list this is the main portion of the window which displays various information about each restricted user. User name . The name of the user that is restricted from hpss access. Realm name . The name of the hpss realm that encompasses the restricted user. User id ....

  • Page 31

    Chapter 3. Using ssm 3.1. The ssm system manager 3.1.1. Starting the ssm system manager before starting the ssm system manager (sm), review the sm key environment variables described in the hpss installation guide, section 3.7.10: storage system management . If the default values are not desired, ov...

  • Page 32

    To help mitigate this, when the thread pool is full, the system manager notifies all the threads in the thread pool that are waiting on list updates to return to the client as if they just timed out as normal. This could be as many as 15 threads per client that are awakened and told to return, which...

  • Page 33

    Port hpssgui and hpssadm clients must access to reach the system manager. This task can be made a bit easier if the system manager rpc program number is labeled in the portmapper. To do this, add a line for the system manager in the /etc/rpc file specifying the program number and a convenient rpc se...

  • Page 34

    · when you have decided on the hpssgui command line that is best for your installation, it will probably be useful to put the command in a shell script for the convenience of all ssm administrators and operators. For example, create a file called “gui” and put the following in it: /opt/hpss/bin/hpss...

  • Page 35

    • the proper authorization entries for the user are created in the authzacl table. 3. The proper ssm configuration files are created and installed. See section 3.3.1: configuring the system manager authentication for ssm clients , section 3.3.2: creating the ssm user accounts , and section 3.3.3: ss...

  • Page 36

    % /opt/hpss/bin/hpssuser -add john -ssm [ adding ssm user ] 1) admin 2) operator choose ssm security level (type a number or return to cancel): > 1 [ ssm user added : admin ] after ssm users are added, removed, or modified, the system manager will automatically discover the change when the user atte...

  • Page 37

    Access to the hpss_server_acl program, hpssuser program, to the hpss db2 database, and to all hpss utility programs should be closely guarded. If an operator had permission to run these tools, he could modify the type of authority granted to anyone by ssm. Note that access to the database by many of...

  • Page 38

    Keytabs are created for the user by the hpssuser utility when the krb5keytab or unixkeytab authentication type is specified. Keytabs may also be created manually with the hpss_krb5_keytab or hpss_unix_keytab utility, as described below. 3.3.2.3.1. Keytabs for kerberos authentication: hpss_krb5_keyta...

  • Page 39

    3.3.3. Ssm configuration file the hpssgui and hpssadm scripts use the ssm configuration file, ssm.Conf for configuration. The mkhpss utility will create the ssm configuration file for the security mechanism supported by ssm. The mkhpss utility will store the generated ssm.Conf at $hpss_path_ssm; the...

  • Page 40

    File option command line option functionality hpss_ssm_sm_host_name -h system manager hostname hpss_ssm_user_pref_path -i path to ssm preferences java_bin -j path to java bin directory krb5_config -k full path to krb5.Conf file hpss_ssm_lang_code -l language code, internationalization ssm_config -m ...

  • Page 41

    File option command line option functionality hpss_authen_type -t authenticator type information on tuning client polling rates for optimal performance is available in the hpssadm and hpssgui man pages. Options are specified, in precedence order, by 1) the command line, 2) the user's environment (se...

  • Page 42

    Note that having encryption types other than "des-cbc-crc" first on the "default_tkt_enctypes" and "default_tgs_enctypes" lines can cause authentication failures. Specifically, keytab files generated by the hpss utility programs will use the first encryption type and only "des-cbc-crc" is known to w...

  • Page 43

    3.3.5.1. Automatic ssm client packaging and installation the hpssuser utility provides a mechanism for packaging all the necessary client files required to execute the hpssgui program on the user's desktop host. Refer to the hpssuser man page for more information on generating an ssm client package....

  • Page 44

    These files may be installed in any location on the ssm client machines. The user must have at least read access to the files. The ssm startup scripts hpssgui.Pl, hpssgui.Vbs, hpssadm.Pl, and hpssadm.Vbs provide the user with a command line mechanism for starting the ssm client. The hpssgui.Pl scrip...

  • Page 45

    3.3.6.2. Solutions for operating through a firewall ssm can operate through a firewall in three different ways: • the hpssgui and hpssadm can use ports exempted by the network administrator as firewall exceptions. See the -n option described in the hpssgui and hpssadm man pages. • the hpssgui and hp...

  • Page 46

    • verify that the proper version of java is installed. Add the java bin directory to the user's $path, or use the -j switch in the hpssgui script, or set java_bin in the user's ssm.Conf file. Java can be downloaded from http://www.Java.Com. • obtain files from the server machine: • obtain the prefer...

  • Page 47

    If access through the firewall is needed for other ports (eg., the kerberos kdc), set up a separate tunnel for each port the firewall does not allow through. • on the client machine, run the gui: • for kerberos authentication: % hpssgui.Pl -s hpssgui.Sessionlog -k krb5.Conf -n 49999 -h localhost • f...

  • Page 48

    Applicable to the platform on which the graphical user interface is running. Custom look and feels are also available at http://www.Javootoo.Com · -b "background color" • the only look and feel that supports color settings and themes is the metal look and feel. The color may be set by using the colo...

  • Page 49

    Include “static” text painted on the window background or labels on things like buttons. Text fields may appear as single or multiple lines and they may be “enterable” (the displayed data can be altered) or “non-enterable” (the displayed data cannot be changed directly). • non-enterable text fields ...

  • Page 50

    • select/cut/copy/paste operations can be performed on enterable text fields; on non-enterable fields, only select and copy operations can be performed. • in some cases, modifying a field value or pressing a button causes the action to be performed immediately. A confirmation window will pop up to i...

  • Page 51

    All of the current configuration’s field values. • freeze - a checkbox that, while checked, suspends the automatic updates made to an ssm window. This allows reviewing information at the frozen point in time. Unchecking the checkbox will reactivate normal update behavior. • refresh button - requests...

  • Page 52

    Menu item is available on all ssm windows. • edit menu - the edit menu is located on all ssm data windows. From each edit menu, the user can access cut , copy and paste functions which enable the user to remove data from text fields or transfer data among them. Editable text fields can be updated. N...

  • Page 53

    Variables may be overridden. 3.8. Monitor, operations and configure menus overview the monitor, operations and configure menus are used by the system manager to monitor, control and configure hpss. They are available only from the hpss health and status window. This section provides a brief descript...

  • Page 54

    Accounting status. Opens the subsystem list window where the accounting status and start accounting buttons can be found. Log files information. Opens the log files information window to display information for the hpss log files such as the log file's size and state. Lookup hpss objects. This subme...

  • Page 55

    Volume labels can be entered and a request to add the disks to a storage class can be submitted. • create tape resources. Opens the create tape resources window where a list of tape volume labels can be entered and a request to add the tapes to a storage class can be submitted. • delete resources. O...

  • Page 56

    Of the accounting policy. Only one accounting policy is allowed. • location. Opens the location policy window allowing configuration and management of the location policy. Only one location policy is allowed. • logging. Opens the logging policies list window allowing configuration and management of ...

  • Page 57

    The hpss login window appears after starting the hpssgui script. The user must supply a valid hpss user name and password in order to access ssm and monitor hpss. If a login attempt is unsuccessful, review the user session log for an indication of the problem. See the hpssadm or hpssgui man pages fo...

  • Page 58

    Mismatched versions may cause compatibility problems. 3.9.2. About hpss the about hpss window displays version information and a portion of the hpss copyright statement. The about hpss window is accessible by selecting the help menu's “about hpss” submenu from any of the hpssgui windows. The hpss sy...

  • Page 59

    When a user successfully connects to the system manager through the login window, the hpss health and status window replaces the login window on the screen. The hpss health and status window will remain on the screen until the user exits or logs out. It provides the main menu and displays informatio...

  • Page 60

    Icon is red, the client’s connection to the system manager is lost; when it is green, the connection is active. 3.9.3.2. Hpss status on the upper section of the hpss health and status window are four status fields that represent the aggregate status of the hpss system. These fields are: servers . Di...

  • Page 61

    In addition to the text which describes the status, these fields are displayed with colored icons. The icon color depicts that status as follows: • red - major and critical problems • magenta – minor problems • yellow - unknown, stale, suspect, and warning problems • green - normal, no problem click...

  • Page 62

    Hpss statistics fields show general trends in hpss operations; the numbers are not all-inclusive. Some values may fluctuate up and down as servers are started or shut down. Some values, such as bytes moved , can be reset to zero in individual movers and by ssm users. Bytes moved . Total bytes moved ...

  • Page 63

    User the ability to hide or display elements of the hpss health and status window in order to optimize the viewable area. Under the view menu there is a menu item and checkbox for each window element that can be hidden. If the box contains a check mark then the corresponding section of the hpss heal...

  • Page 64

    Field descriptions start time . The time the system manager was started. Uptime . The elapsed clock time since the system manager was started. Hpss management guide november 2009 release 7.3 (revision 1.0) 64.

  • Page 65

    Cpu time. The amount of cpu time that the system manager has consumed. Memory usage. The amount of memory that the system manager is currently occupying. Process id. The process id of the system manager. Hostname. The name of the host where the system manager is running. Rpc calls to servers. The nu...

  • Page 66

    Thread pool size and/or request queue size to help with the system manager performance. However, increasing these 2 parameters could cause the system manager to require more memory. • data change notifications. The number of data change notifications received from servers. (server rpc interface only...

  • Page 67

    • hostname. The name of the host where the client is running. • connections. The number of rpc connections this client has to the system manager. • start time. The time that the client connected to the system manager. • connect time. The elapsed time since the client connected to the system manager....

  • Page 68

    Field descriptions user login name . The name that the user entered as the user id when logging into hpss. User authority. The authority level that the user has in order to perform operations in ssm. This can be admin or operator. Login time. The time that the user logged into ssm. Total session tim...

  • Page 69

    Percent memory free . The ratio of free memory to total memory in the hpssgui process. Total windows opened during this session . The number of windows created during the current user session. 3.10. Ssm list preferences when a user logs into ssm, the system manager reads the saved preferences file a...

  • Page 70

    Hpssgui_user_cfg_path or the configuration file entry hpss_ssm_user_pref_path. If this option is not specified, the default value is :/hpss-ssm-prefs . The user must have permissions to create the preferences file in the directory. Preferences windows contain filters for controlling the data display...

  • Page 71

    Save . Save the current preference settings to the preferences file using the same preference name. Delete . Delete the currently displayed preference settings from the preferences file. Sick and default preference settings cannot be deleted. Reload . Reread the preference settings from the preferen...

  • Page 72

    Chapter 4. Global & subsystem configuration this chapter discusses two levels of system configuration: global and storage subsystem. The global configuration applies to the entire hpss installation while subsystem configurations apply only to servers and resources allocated to storage subsystems. Fo...

  • Page 73

    This window allows you to manage the hpss global configuration record. Only one such record is permitted per hpss installation. To open the window, on the health and status window select the configure menu, and from there the global menu item. Field descriptions system name. An ascii text string rep...

  • Page 74

    Storage subsystem configuration. Root user id. The uid of the user who has root access privileges to the hpss namespace. This only applies if the root is superuser flag is set. Cos change stream count. The number of background threads that run in the core server to process class of service change re...

  • Page 75

    This window lists all the subsystems in the hpss system and provides the ability to manage these subsystems. To open the window, from the health and status window select the configure menu, and from there select the subsystems menu item. To create a new subsystem, click on the create new button. To ...

  • Page 76

    Subsystem is selected in the list. Delete - delete the selected subsystem(s). This button is disabled unless a subsystem is selected in the list. Always contact hpss customer support before deleting a subsystem definition. An improperly deleted subsystem can cause serious problems for an hpss system...

  • Page 77

    This window allows an administrator to manage the configuration of a storage subsystem. The add button is only displayed during the creation of a new configuration. The update button is displayed when an existing configuration is being modified. To open this window for creation of a new subsystem, c...

  • Page 78

    User does not specify a cos or any hints with the creation request. The global configuration specifies a default cos for an entire hpss installation. Selecting a cos on the storage subsystem configuration window allows the global value to be overridden for a particular subsystem. If the field is bla...

  • Page 79

    Db log monitor interval. The core server will check consistency of database logs and backup logs at the indicated interval, specified in seconds. The logs are consistent if both primary and backup log directories exist and contain log files with the same names. The minimum value for this field is 30...

  • Page 80

    Gatekeeper is already configured, simply add it to your storage subsystem's configuration. However, if it is not yet configured, it will be necessary to wait until section 4.2.3.4: assign a gatekeeper if required on page 80 to add the gatekeeper. 6. Set the metadata space thresholds and the update i...

  • Page 81

    4.2.3.7. Migration and purge policy overrides the migration and purge policies contain two elements, the basic policy and the storage subsystem specific policies. This can be seen on the migration policy and purge policy windows. If a given migration or purge policy does not contain any subsystem sp...

  • Page 82

    E. Issue the following sql command: db2> select count(*) from nsobject the result of the command should indicate 2 rows in this table. 3. If any of these checks gives an unexpected result, do not delete the subsystem. Contact hpss customer support. When deleting an existing storage subsystem, it is ...

  • Page 83

    Chapter 5. Hpss servers most hpss server administration is performed from the ssm graphical user interface servers list window . Each hpss server has an entry in this list. 5.1. Server list this window facilitates management of the configured hpss servers. From this window, an hpss server can be sta...

  • Page 84

    The server’s configuration should be carefully reviewed to ensure that it is correct and complete. Check the alarms and events window and the hpss log file to view ssm alarm messages related to configuration problems. This situation can be caused by: • a db2 record required by the server is missing ...

  • Page 85

    Rate (see the hpssgui/hpssadm man pages for more details). If a server is configured as executable but is not running, ssm will treat it as an error. Therefore, if a server is not intended to run for an extended period, its executable flag should be unchecked. Ssm will stop monitoring the server and...

  • Page 86

    Execute host . The execute hostname field from the server's basic configuration record. This field is intended to specify the hostname on which the server is supposed to run; however, no checking is done to verify if the server is actually running on the specified host. This field is only used by th...

  • Page 87

    Server hangs up or otherwise won't respond to the shutdown command. Force connect - request the system manager to immediately attempt to connect to the selected servers. The system manager routinely attempts to connect to any unconnected servers; using this button will simply cause the next attempt ...

  • Page 88

    • storage system manager • startup daemon (on each host where an hpss server will be executing) the fields of the server configuration window are divided into the following sections. The basic controls section is at the top of the window and the other sections are on individual tabs: • basic control...

  • Page 89

    ● section 5.1.3: mover specific configuration on page 102 ● section 5.1.1: physical volume repository (pvr) specific configuration on page 109 ● details about all of the other sections on this window, which apply to all server types, are described in section 5.1.1: common server configuration . To v...

  • Page 90

    Field descriptions server name. A unique descriptive name given to the server. Ensure that the server name is unique. A server’s descriptive name should be meaningful to local site administrators and operators, in contrast to the server’s corresponding uuid, which has meaning for hpss. For hpss syst...

  • Page 91

    Execute hostname. This is the hostname of the node on which the server will execute. It must match the execute hostname of the startup daemon that is to manage this server. For most servers, setting this field is straightforward, but for remote movers, this indicates the node on which the mover admi...

  • Page 92

    5.1.1.2. Interface controls the interface controls section of the server configuration window is common to all servers. In the example window above, the server displayed is a core server. Field descriptions maximum connections. The maximum number of clients that this server can service at one time. ...

  • Page 93

    The security controls section of the server configuration window is common to all servers. In the example window above, the server displayed is a core server. Field descriptions principal name. The name of the principal the server will use to authenticate. Protection level. The level of protection t...

  • Page 94

    Authenticator. The argument passed to the authentication mechanism indicated by the authenticator type configuration variable and used to validate communications. If it is a keytab, the server must have read access to the keytab file. Other access permissions should not be set on this file or securi...

  • Page 95

    • utime. Core server bitfile time modified events. • acl_set. Core server access control list modification events. • chbfid. Core server change bitfile identifier events. • bfsetattrs. Core server set bitfile attribute events. 5.1.1.1. Log policy the server log policy may also be accessed from the l...

  • Page 96

    • trace. If selected, trace messages generated by the server are sent to the log. It is recommended that this be off for all servers except the mover. These messages give detailed information about program flow and are generally of interest only to the server developer. In normal operation, logging ...

  • Page 97

    Cos change retry limit , tape dismount delay , tape handoff delay , pvl max connection wait , fragment trim limit and fragment smallest block can be changed in the core server while the server is running by changing the value on this screen, updating the metadata, then re-initializing the appropriat...

  • Page 98

    The server uses built-in default values for these settings, but if the environment variables can be found in the server's environment, the server uses those values. The following is a list of the names of the variables and the aspects of the server's operation they control. Since the environment is ...

  • Page 99

    For both gatekeeping and account validation. If multiple gatekeepers are configured, then any gatekeeper may be contacted for account validation requests. Note: if a gatekeeper is configured, then it will either need to be running or marked non-executable for hpss client api requests to succeed in t...

  • Page 100

    5.1.4. Log client specific configuration this window controls the local log settings that will be in effect for the node on which this log client runs. Field description client port. The port number for communication between the log client and the hpss servers. The default value is 8101. Ensure that...

  • Page 101

    Related information section 9.5: managing local logging on page 301 . 5.1.1. Log daemon specific configuration this window controls configuration of the log daemon and the central hpss log. Field descriptions log file maximum size. The maximum size in bytes of each central log file. The default valu...

  • Page 102

    Field descriptions storage class update interval (seconds). The interval that indicates how often the mps will query the core server in its subsystem to get the latest storage class statistics. This is also the interval the mps uses to check whether it needs to initiate a purge operation on the stor...

  • Page 103

    And usage. The trade-off for this value is that large buffer sizes will use more system memory and may be inefficient for small transfers (e.G., if the mover buffer size is 4mb, but client requests are 512kb, the mover will not achieve any double buffering benefit because the entire amount of the tr...

  • Page 104

    Range end. If this field is zero, port range end field must also be zero. Port range end. Used in conjunction with port range start (see above). Valid values are zero or any tcp port number to which the mover may bind (that is greater than or equal to the value of port range start ). The default val...

  • Page 105

    This will cause inetd to run the executable /opt/hpss/bin/hpss_mvr_tcp under the root user id when a connection is detected on port 5002. The mover process uses the /var/hpss/etc/mvr_ek file to read the encryption key that will be used to authenticate all connections made to this mover. After modify...

  • Page 106

    /etc/xinetd.D file. For example, if the encryption key in the mover’s type specific configuration is 1234567890abcdef , then the encryption key file (/var/hpss/etc/ek.Mvr1) should contain: 0x12345678 0x90abcdef 5.1.3.1.3. /var/hpss/etc files required for remote mover the mover process on the remote ...

  • Page 107

    Table 1. Irix system parameters parameter name minimum value parameter description semmsl 512 maximum number of semaphores per set maxdmasz 513 maximum dma size (required for ampex dst support) solaris solaris system parameters which affect the remote mover can be modified by editing the /etc/system...

  • Page 108

    Table 3. Linux system parameters parameter name minimum value parameter description semmsl 512 maximum number of semaphores per id shmmax 0x2000000 maximum shared memory segment size (bytes) 5.1.3.1.1. Setting up remote movers with mkhpss the mkhpss utility may be used to copy the files needed for a...

  • Page 109

    To all nodes or unexpected results may occur. • the mover must be built with the lft option. This is the default option for all movers. If not all movers have been built with this option, clients must explicitly specify a class of service which is valid for a mover supporting the local file transfer...

  • Page 110

    Specific configuration window on page 120, section 5.1.1.1: 3494 pvr specific configuration on page 111, section 5.1.1.2: aml pvr specific configuration on page 113, and section 5.1.1.2: scsi pvr specific configuration window on page 118). These sections provide additional vendor-specific advice on ...

  • Page 111

    5.1.1.1. 3494 pvr specific configuration 5.1.1.1.1. 3494 pvr specific configuration window field descriptions cartridge capacity . The total number of cartridge slots in the library dedicated to this hpss pvr. This may or may not be the total cartridge capacity of the library; a site might use part ...

  • Page 112

    Only if the support shelf tape checkbox is selected. The alarm value must be 2 or greater. Dismount delay. When defer dismounts is checked, this value is used by the pvl to determine the number of minutes that dismounts are delayed after the last data access. Retry mount time limit. The default valu...

  • Page 113

    Named /dev/lmcp0 and /dev/lmcp1 respectively. Control connections must be made prior to configuration of the /dev/lmcpx devices or undefined errors may result. For linux systems, the symbolic library name defined in /etc/ibmatl.Conf (e.G., 3494a) should be used. For rs-232 and ethernet connected rob...

  • Page 114

    Weight 2 * cartridges from other jobs mounted on this drive’s controller + weight 3 * units of distance from the cartridge to the drive this method has the effect of distributing a striped tape mount across as many controllers as possible for the best performance. It also will try to pick controller...

  • Page 115

    Requesting client to the aml’s storage positions, drives, and insert/eject units. Access configurations for clients are set in the configuration file c:\das\etc\config on the os/2 pc. The client name can be up to 64 alphanumeric characters in length and is case sensitive. Server name. Tcp/ip host na...

  • Page 116

    1. Make sure the amu archive management software is running and the hostname is resolved, 2. Select an os/2 window from the desktop and change the directory to c:\das, c:> cd \das 3. At the prompt, type tcpstart and make sure that tcp/ip gets configured and that the port mapper program is started, c...

  • Page 117

    Score = weight 1 * cartridges from this job mounted on this drive’s controller + weight 2 * cartridges from other jobs mounted on this drive’s controller + weight 3 * units of distance from the cartridge to the drive this method has the effect of distributing a striped tape mount across as many cont...

  • Page 118

    • support shelf tape. If on, the pvr and the pvl will support the removal of cartridges from the tape library using the shelf_tape utility. Command device. The name of the device that the pvr can use to send commands to the robot. For aix systems, this is generally /dev/smc0 . For linux systems, use...

  • Page 119

    Score = weight 1 * cartridges from this job mounted on this drive’s controller + weight 2 * cartridges from other jobs mounted on this drive’s controller + weight 3 * units of distance from the cartridge to the drive this method has the effect of distributing a striped tape mount across as many cont...

  • Page 120

    • enforce home location. If on, the scsi pvr will always try to dismount a mounted cart back to its home location. Otherwise, it will just use the first free slot. The scsi_home utility can be used to view and manipulate the home location values. Serial number. The serial number of the robot, obtain...

  • Page 121

    Requested shelf tape has been checked-in. The pvr will continue checking at this interval until the tape is checked-in. This field applies only if the support shelf tape checkbox is selected. The retry value must be 30 or greater. Shelf tape check-in alarm . The pvr will periodically log alarm messa...

  • Page 122

    Hpss will use any cartridge access port (cap) in the stk robot that has a priority greater than zero. When it needs a cap, hpss will pick the highest priority cap that is currently available. At least one cap must be assigned a non-zero priority. See the stk automated cartridge system library softwa...

  • Page 123

    Would you like multi-host or single-host testing? Enter one of the following followed by enter: m multi-host testing s single-host testing x exit this script enter choice: m would you like to define the server side or client side for multi-host testing? Enter one of the following followed by enter: ...

  • Page 124

    Server are removed from the hpss configuration. The steps described in this section are general guidelines. Specific procedures should be worked out with the aid of hpss technical support so that the details of the system's configuration can be considered. A server’s configuration should be removed ...

  • Page 125

    Pressing the delete button. 5.1. Monitoring server information a server that is running and connected to ssm will allow the ssm user to view and update its information. This section describes the server execution statuses and configuration information. A typical hpss server allows the ssm users to c...

  • Page 126

    • busy - the server is busy performing its function. Most servers do not update usage state dynamically, so it is unlikely you will see this value reported. • unknown - the server has not reported a recognized usage state. Administrative state . The administrative state of the server. The possible s...

  • Page 127

    Communication status: normal however, when the server is experiencing errors or encountering abnormal conditions, it will change the appropriate states and statuses to error values, notify ssm of the changes, and issue an alarm to ssm. Refer to section 9.6.2: alarm/event information on page 303 for ...

  • Page 128

    Global database name . The name of the global database for the hpss system. Subsystem database name . The name of the database which contains the subsystem tables used by this core server. Schema name . The name of the database schema. Root fileset name . The name of the root fileset used by the cor...

  • Page 129

    • file deletes. The number of bitfile delete requests processed in the core server since startup or last reset of the statistics. • last reset time. The last time the subsystem statistics were reset. If this value is 0, the statistics have not been reset since server startup. Name space statistics: ...

  • Page 130

    • free tape bytes . This is an estimate based on the sum of the estimated sizes of the partially written and unwritten tape volumes. It is not, and cannot be, an accurate value as the amount of data that can be written on tapes varies with individual tape volumes and data compression levels. Options...

  • Page 131

    Must be greater than zero and is used if the gatekeeping site interface returns a wait time of zero for the create, open, or stage request being retried. Changing the value of this field will cause the gatekeeper to use the new value until the next restart at which point it will then go back to usin...

  • Page 132

    Server when the gatekeeper is monitoring requests and a client disconnects. • get monitor types . Statistics from the gk_getmonitortypes api. This api is called by the core server to figure out what types of requests being monitored by the gatekeeper. • pass thrus . Statistics from the gk_passthru a...

  • Page 133

    Minimum location map update time. The shortest time, in seconds, needed for a location map update. Average location map update time. The average time, in seconds, needed for a location map update. Maximum location map update time. The longest time, in seconds, needed for a location map update. Assoc...

  • Page 134

    Changes made to fields on this window are sent directly to the mover after the appropriate button is pressed and are effective immediately. Field descriptions server name. The descriptive name of the mover. Number of request tasks. The number of mover request-processing tasks that currently exist. N...

  • Page 135

    This window allows you to view the type-specific information associated with a pvl. Field descriptions server name . The descriptive name of the pvl. Total volumes . The total number of volumes that have been imported into the pvl. Total repositories . The total number of pvrs in the servers list wi...

  • Page 136

    Characteristics. Flags for the pvr: • defer dismounts. If on, the pvl will delay the dismounting of a tape cartridge until the drive is required by another job or until the dismount delay time limit is exceeded. • support shelf tape. If on, the pvr and the pvl will support the removal of cartridges ...

  • Page 137

    Heavily used controller, then a more distant drive will be selected. Retry mount time limit. The default value for this field is -1. When the default value (-1) is used, if an error is encountered during a pvr mount operation, the mount will pend and be retried every 5 minutes. Setting a value in th...

  • Page 138

    5.1.1.2.1. Aml pvr information window field descriptions server name . The descriptive name of the pvr. Total cartridges . The number of cartridges currently being managed by the pvr. Cartridge capacity . The total number of cartridge slots in the library dedicated to this hpss pvr. This may or may ...

  • Page 139

    The number of consecutive mount errors which occur to any drive in this pvr equal or exceed this value, the drive is automatically locked by the pvl. The only mount errors that apply are those set through the retry mount time limit mechanism. The drive error count field in the pvl drive information ...

  • Page 140

    Alarm. Same job on controller , other job on controller , & distance to drive . These values are used by the pvr when selecting a drive for a tape mount operation. The three values are essentially weights that are used to compute an overall score for each possible drive. After the score has been cal...

  • Page 141

    Shelf tape check-in retry . The number of seconds the pvr will wait before asking the robot if a requested shelf tape has been checked-in. The pvr will continue checking at this interval until the tape is checked-in. This field applies only if the support shelf tape checkbox is selected. The retry v...

  • Page 142

    The best performance. It also will try to pick controllers that are currently driving a minimum number of tapes. So, in an environment with many tape drives per controller, the above algorithm will minimize the load on any one controller. The distance to drive helps minimize mount times by mounting ...

  • Page 143

    5.1.1.2.1. Stk pvr information window field descriptions server name . The descriptive name of the pvr. Total cartridges . The number of cartridges currently being managed by the pvr. Cartridge capacity . The total number of cartridge slots in the library dedicated to this hpss pvr. This may or may ...

  • Page 144

    Server to set the vv condition of the associated tape volume to down. Once in down state, the volume will no longer be available for read or write operations. For further information about the core server vv condition, see section 4.5.4.2: core server tape volume information window on page 271. Driv...

  • Page 145

    Hpss administrators and operators may use ssm to view the active rtm requests. The rtm summary window lists a summary of the current rtm w requests. The rtm detail window displays detailed information for selected i rtm requests. 5.1.1. Rtm summary list field descriptions rtm summary list . This is ...

  • Page 146

    5.1.2. Rtm detail the rtm detail window displays a snapshot of the details of the selected rtm requests from the rtm summary list window. This may contain information from multiple servers, gatekeeper, core and mover. The actual data displayed will be different for each server type and is displayed ...

  • Page 147

    Reqid. The rtm request identifier. Reqcode. The action or operation that this request is currently executing. Examples include "mover write", "pvl verify", "write tm" (tape mark), etc. Reqstate. The state of the requested operation. Examples include "in progress", "suspended", "blocked", etc. Server...

  • Page 148

    Pvljobid. The id of the pvl job associated with this request. Mvrid. The id of the mover this request is currently waiting on. Deviceid. The device this request is curretly waiting on to complete a data move operation. Segment. The id of the storage segment being operated on by this request. Vv. The...

  • Page 149

    Groupid. The group id of the user associated with this request. Hostaddr. The address of the originating host requesttype. The type of this request (open, create or stage). Oflag. The open flags associated with this file open request. Stageflags. Flags associated with this file stage operation. Stag...

  • Page 150

    To start the startup daemon, use the “-d” option to rc.Hpss: % su - % /opt/hpss/bin/rc.Hpss -d [start] 5.2.2.2. Starting ssm the ssm system manager configuration metadata should have already been created by mkhpss as part of the the infrastructure configuration. After ssm is started, this configurat...

  • Page 151

    To start a server, select the desired server(s) from the servers window and click on the start button. Verify the result of the request in the message area on the servers window. In addition, monitor the alarms and events window for the “server initialized” event. Reference section 5.2.2.3: on page ...

  • Page 152

    Window for the “server terminated” event. The hpss startup daemon(s) and the ssm system manager cannot be shut down from the servers window. Select system manager from the shutdown submenu of the operations menu of the health and status window to shut down the system manager. Use the rc.Hpss script ...

  • Page 153

    Command line sessions, will detect that the system manager has exited. Choosing this option will pop up a confirmation window which allows the shutdown request to be approved or canceled. As the system manager exits, a notification window will pop up on each logged on ssm graphical user interface se...

  • Page 154

    Button. Verify the result of the request in the status bar on the servers window. In addition, monitor the alarms and events window for the “server repaired” event. Repairing a server does not correct the underlying problem that caused the server's reported state to change. Rather, it is a means for...

  • Page 155

    Core server resets the cos copy to disk , cos change retry limit , tape dismount delay , tape handoff delay , pvl max connection wait , fragment trim limit and fragment smallest block values to the values in the specific configuration metadata record. Reloads cached class of service information for ...

  • Page 156

    Servers that do not support reinitialization, or those that do not support reinitializing the settings in question, must be restarted in order for configuration modifications to take affect. Some groups of servers depend on consistent configuration information to run properly. For example, the core ...

  • Page 157

    Chapter 6. Storage configuration this chapter describes the procedures for creating, modifying, and deleting storage classes, hierarchies, classes of service, migration policies, purge policies, and file families. 6.1. Storage classes this section describes the process of configuring storage classes...

  • Page 158

    Information buttons. Migration policy. Opens the configuration window for the migration policy that is configured for the selected storage class. This button will be disabled if no storage classes are selected in the storage classes list or the selected storage class does not have a migration policy...

  • Page 159

    This window is used to manage disk storage class configurations. Field descriptions storage class id. The numeric identifier assigned to the storage class. Storage class name. The descriptive name of the storage class. Storage class type. The type of the storage class (disk). Migration policy. The m...

  • Page 160

    Advice - do not configure a migration policy for a storage class at the lowest level in a hierarchy. If a migration policy is added to a storage class after files are created in the storage class, those files may never be migrated. Use the mkmprec utility to correct this problem. See the mkmprec man...

  • Page 161

    Vvs, fragmentation of the volumes may make it difficult to find space for a new segment. Setting average number of storage segments to a larger value will increase the number of segments occupied by files, and decrease the segment size. Fragmentation of the volumes will be reduced, but the amount of...

  • Page 162

    Min storage segment size (minseg). The lower bound for storage segment sizes created on volumes in this storage class. This value is the product of the stripe length (sl) and the min multiplier (minmult) . Max multiplier (maxmult). The max storage segment size (maxseg) must be a power of 2 multiple ...

  • Page 163

    This window is used to manage tape storage class configurations. Field descriptions storage class id. The numeric identifier assigned to the storage class. Storage class name. The descriptive name of the storage class. Storage class type. The type of the storage class (tape). Migration policy. The m...

  • Page 164

    If a migration policy is added to a storage class after files are created in the storage class, those files may never be migrated. Use the mkmprec utility to correct this problem. See the mkmprec man page for more information. Warning threshold. A threshold for space used in this storage class expre...

  • Page 165

    Movement protocol overhead and helps to keep the data streams flowing smoothly. Vv block size must meet the following constraining requirements: • it must be an integer multiple of the media block size. • the pv section length ( media block size (mbs) * blocks between tape marks (bbtm ) divided by t...

  • Page 166

    If the tape media supports "fast locate", and that feature is enabled for the tape devices, choose larger values of seconds between tape marks (sbtm) . When reading from the middle of a file on tape, the fast locate feature is used by hpss to locate the data block in which a given portion of a file ...

  • Page 167

    This window is used to define warning and critical thresholds unique to a particular storage subsystem, overriding the values defined in the disk storage class. The user may modify either the warning percent, critical percent or both for one or more of the listed subsystems. Select a subsystem from ...

  • Page 168

    Subsys name. The name of the selected storage subsystem. Warning. The current warning threshold value for the selected subsystem. If the storage class defaults are to be used, the text “default” will be displayed. Critical. The current critical threshold value for the selected subsystem. If the stor...

  • Page 169

    Warning volumes, critical volumes or both for one or more of the listed subsystems. Select a subsystem from the list and then modify the values on the lower portion of the window. When the new values have been entered, select update to commit the changes. To remove the customized threshold values, s...

  • Page 170

    The administrator to reflect desired behavior. If files have been stored in a storage class without a migration policy, and a migration policy is subsequently configured for it, the files created before the addition of the policy will not be migrated. Use the mkmprec utility to create migration reco...

  • Page 171

    Accessed from the hpss health and status window's configure menu, submenu storage space , item heirarchies. Refer to section 3.9.3: hpss health and status on page 58. The following rules for creating storage hierarchies are enforced by the hierarchies window: • a storage class may be used only once ...

  • Page 172

    Configuration buttons create new. Open a storage hierarchy window with default values. Configure. Open the selected storage hierarchy configuration for editing. One hierarchy from the list must be selected before this button is active. Delete. Delete the selected storage hierarchy(s). 6.2.2. Storage...

  • Page 173

    Field descriptions hierarchy id. The id associated with this hierarchy . Any unique, positive 32-bit integer value. The default value is the last configured id plus 1. Hierarchy name. The descriptive name associated with this hierarchy. The default value is “hierarchy ”. Top storage class. The stora...

  • Page 174

    Do this will render the files unreadable. It is recommended that hpss customer support be called to assist with this operation. 6.3. Classes of service this section describes how to configure classes of service. 6.3.1. Classes of service window a cos can be created and managed using the classes of s...

  • Page 175

    Create new. Open a class of service window containing default values for a new class of service. Configure. Open the selected class(es) of service configuration(s) for editing. Delete. Delete the selected class(es) of service. 6.3.2. Class of service configuration window this window allows an admini...

  • Page 176

    Is on open. For all subsequently created coses, the default value is the same as the most recent cos configured. Advice – changing the stage code should be done with care. See section 6.3.3: changing a class of service definition on page 178 for detailed information on each of the choices for stage ...

  • Page 177

    • auto stage retry . When this flag is turned on, and a valid secondary copy of the data exists, and a stage from the primary copy fails, hpss will automatically retry the stage using the secondary copy. • auto read retry. When this flag is turned on, and a valid secondary copy of the data exists, a...

  • Page 178

    Users do frequent appends, or if those who do can be relied upon to turn truncation off for their own files, or if the system administrator can easily identify files which are frequently appended and can turn off truncation on them individually, then the site might want to take advantage of the spac...

  • Page 179

    Have a significant impact. Turning the flag on constrains files that are already larger than the maximum file size to their current size. Existing smaller files will be constrained to the maximum file size . Changing minimum file size can have an impact on cos selection. Currently, the pftp and ftp ...

  • Page 180

    6.3.5. Changing a file's class of service the core server provides a means to change the class of service of a file. The core server moves the body of the file as appropriate to media in the destination class of service, then allows the usual migration and purge algorithms for the new class of servi...

  • Page 181

    Both basic and subsystem specific migration policies are created and managed using the migration policies window. The basic policy must be created before creating any subsystem specific policies. The fields in the basic policy are displayed with default values. Change any fields to desired values as...

  • Page 182

    Other migration policy list columns. The remaining columns provide the same information that can be found in section 6.4.2.1: disk migration policy configuration on page 182 and section 6.4.2.2: tape migration policy configuration on page 185 windows. Configuration buttons. Create disk. Opens a disk...

  • Page 183

    This window allows an administrator to manage disk migration policies and their subsystem-specific overrides. Subsystem-specific policies define migration rules to be applied on a subsystem basis instead of using the default migration policy. When a migration run occurs and the storage class’s migra...

  • Page 184

    Last update interval. The number of minutes that must pass since a file was last updated before it can become a candidate for migration. Number of migration streams per file family. The number of migration streams to be allocated to each file family. This value effectively determines how many file f...

  • Page 185

    Storage class. Triggered migration options. There are four choices for managing migration behavior when a storage class is running out of space and the next migration isn’t yet scheduled to occur: migrate at warning threshold. A migration run should be started immediately when the storage class warn...

  • Page 186

    This window allows an administrator to manage tape migration policies and their subsystem-specific overrides. Subsystem-specific policies define migration rules to be applied on a subsystem basis instead of using the default (basic) migration policy. When a migration run occurs and the storage class...

  • Page 187

    Migration criteria. This goal may not be attainable if the total size of all files not eligible for migration is large. Total migration streams. This value determines the degree of parallelism in the file migration process. This applies to policies using the migrate files and migrate files and purge...

  • Page 188

    Based on individual files rather than tape volumes, and is able to make second copies of files stored on tape. In this algorithm, individual files are selected for migration based on their last write time and the settings in the migration policy. The selected files are migrated downwards to the next...

  • Page 189

    To delete a migration policy, select the policy from the migration policies list window and press the delete button. If a basic policy is selected, and the policy has sub-system specific policies associated with it, a prompt will appear asking if the basic policy and the related sub-system specific ...

  • Page 190

    Policy button after the window refreshes, enter the specific purge policy parameters, and press the update button. This process can be repeated for each sub-system. When a purge policy is added to an existing storage class, the migration purge servers must be restarted in order for the policy to tak...

  • Page 191

    This window allows you to manage a purge policy. Purge policies are assigned to storage classes to tell the migration purge server and core server how to free disk space occupied by files which have been migrated. Purge policies apply to disk storage classes only. The window always includes the basi...

  • Page 192

    Unaccessed (for read or write) for the length of time specified by this field. Start purge when space used exceeds. Purge will begin for a storage class when the amount of its space used exceeds this threshold. Used space includes any file in the storage class, whether it has been migrated or not. S...

  • Page 193

    Policy are applied to all storage classes which reference the policy. If the policy is reread, the changes are only applied to the storage class and storage subsystem for which the policy is reread. Core servers are not able to reread purge policies. If the purge by field is changed on either a basi...

  • Page 194

    Field descriptions the fields of the columns of this window are those of the file family configuration described in section 6.6.1: file family configuration . Configuration buttons create new. Open a file family configuration window with default values. Configure. Open the selected file family confi...

  • Page 195

    Hpss management guide november 2009 release 7.3 (revision 1.0) 195.

  • Page 196

    Chapter 7. Device and drive management every disk and tape drive that is used by hpss is controlled by two servers. The pvl controls mounts and dismounts (for disk devices these are logical operations only), and the mover controls i/o. In support of these two views, the terms “pvl drive” and “mover ...

  • Page 197

    May result in a large performance degradation. For san disk devices, the hpss_san3p_part utility with the -i option and device name must be run to assign a uuid to the disk device. This uuid should be used in the device name field when configuring the disk device. Currently the number of drives whic...

  • Page 198

    These windows allow you to manage a tape or disk device/drive configuration. Modifying/updating the device/drive configuration via the tape device configuration or disk device configuration windows is not permitted while the pvl, the associated mover(s), or the associated pvr(s) (for tape) are runni...

  • Page 199

    For irix systems, scsi attached tape drives are typically referred to by pathnames of the form /dev/rmt/ tpsxdyns, where x is the scsi controller number, and y is the scsi id of the drive. Note that for ampex dst drives, the tpsxdynrns name should be used (indicating that the driver should not attem...

  • Page 200

    Advice - this option is supported for 3590, 3590e, 3580, 3592, 9840, 9940, dst-312, dst-314, t10000 and gy-8240 devices. • no-delay support (tape only) . An indication of whether the device supports opening the device with no delay flag set, while allowing tape i/o operation after the open. Advice -...

  • Page 201

    Device. Without the reservation, it is possible for other hosts to interleave scsi commands to the drive with those issued by hpss. This effect could potentially lead to corruption of data. Table 2. Recommended settings for tape devices device driver mover executable no-delay support locate support ...

  • Page 202

    Must be a valid id. The valid ids can be found in the affinity list for any cartridge in the robot. Use the command “mtlib -l -qv -v” to obtain the affinity list for a cartridge. Polling interval (tape only). The number of seconds to wait between polling requests performed by the pvl to determine if...

  • Page 203

    This window allows you to view the list of configured mover devices and pvl drives. It also provides a number of function buttons, which allow certain operations to be performed on devices or drives. The device and drive list preferences window may be used to select the device/drives which will be d...

  • Page 204

    • disabled - the device is locked, which makes it unavailable for use by hpss. • unknown - the state of the device is not known to ssm; this is usually caused by the controlling mover being down or disconnected from ssm. Device admin state . The current administrative state of the device, as reporte...

  • Page 205

    • unknown - the state of the drive is not known to ssm; this is usually caused by the pvl being down or disconnected from ssm comment. This field provides a 128 character buffer in the pvl drive metadata which gives the administrator the opportunity to associate miscellaneous text with a device/driv...

  • Page 206

    Drive administration buttons this group of buttons affects selected drives. All the buttons are disabled unless one or more drives are selected (see figure above). Lock . Lock the selected drives, making them unavailable to hpss. When a drive is locked, the pvl will no longer schedule the pvl drive....

  • Page 207

    Window in add mode, allowing you to create a new disk device and drive. Create tape . This button is always active. Clicking on it opens the tape device configuration window in add mode, allowing you to create a new tape device and drive. Configure. This button is active if one or more devices/drive...

  • Page 208

    Some pvl drive configuration attributes can be updated dynamically using the pvl drive information window (section 7.2.2: pvl drive information window on page 214). The settable fields in this window are updated dynamically (i.E. Saved to metadata and used by the pvl upon successful update). It is a...

  • Page 209

    There are a number of situations in which the pvl won't allow the device/drive to be deleted: • if the device/drive is still attempting to notify the mover/pvr about being added • if the device/drive is in the process of aborting mover/pvr notification • if the drive is in use by the pvl • if the dr...

  • Page 210

    Hpss management guide november 2009 release 7.3 (revision 1.0) 210

  • Page 211

    The mover device information window reports the current statistics for the device, such as the workload history of the device since the startup of the controlling mover. The mover device information window can also be used to lock and unlock a mover device (note: locking the mover device generally i...

  • Page 212

    Means that the corresponding flag is set. • read enabled . An indication of whether the device is available for reading. • write enabled . An indication of whether the device is available for writing. • locate support (tape only). An indication of whether the device supports a high speed (absolute) ...

  • Page 213

    • multiple mover tasks (disk only). If on, the mover will allow multiple mover tasks to access the disk device. • reserve/release (tape only). An indication of whether a scsi reservation is taken on the device when it's opened. Advice - this is useful on fibre attached tape devices to ensure that hp...

  • Page 214

    7.2.2. Pvl drive information window hpss management guide november 2009 release 7.3 (revision 1.0) 214.

  • Page 215

    Hpss management guide november 2009 release 7.3 (revision 1.0) 215.

  • Page 216

    This window allows you to view/update the information associated with an hpss drive. The pvl drive information window is typically used to lock and unlock drives since newly configured drives are locked by default and must be unlocked to be used. It may also be used to determine which volume is moun...

  • Page 217

    Pvr (tape only) . The descriptive name of the pvr used to control this drive. This field is only meaningful for tape drives. Administrative state . This field allows you to modify the state of the drive. The options are: • locked - makes the drive unavailable for hpss requests. • unlocked - makes th...

  • Page 218

    • pvr/mover notify pending - the pvl needs to notify the associated pvr and mover that the drive has been created or deleted. • pvr notify pending - the pvl needs to notify the associated pvr that the drive has been created or deleted. • mover notify pending - the pvl needs to notify the associated ...

  • Page 219

    And thus availability. To do this, the hpss administrator will need to: • associate tape drives to a specific drive pool by configuring the hpss tape drives with a non-zero positive integer drive pool id . • modify the end client to dictate that their read request be serviced by tape drives from thi...

  • Page 220

    7.4. Changing device and drive state the administrative state of a device or drive can be set to unlocked or locked . This controls whether hpss can access the drive. Changing the state of a device or drive can be accomplished via the devices and drives list window. Notice that there are two sets of...

  • Page 221

    Locking a disk drive has little effect since disks are logically mounted when the pvl initializes and are not usually unmounted; however, a disk drive must be in the locked state to be deleted. 7.4.3. Repairing the state of a device or drive a drive can enter an error or suspect state as reported by...

  • Page 222

    Hpss management guide november 2009 release 7.3 (revision 1.0) 222.

  • Page 223

    Chapter 8. Volume and storage management this chapter describes the procedures for adding, removing, monitoring, and managing storage space in the hpss system. The basic unit of storage which can be added to the hpss system is the volume. Before volumes can be added to hpss, the underlying configura...

  • Page 224

    /var/hpss/etc/aml_ejectport.Conf and /var/hpss/etc/aml_insertport.Conf. The aml robot can have multiple insert and eject ports, which have the capability to handle different media types. These two configuration files in conjunction with the amu ams configuration files specify which insert/eject area...

  • Page 225

    Information contained in the internal label will not match the side information passed to pvl in the import request. If the start of the ownerid field is not “hpss”, then the volume will be imported as a foreign label volume and the side will be set to zero. When importing a non-removable disk volum...

  • Page 226

    This window allows the user to import tape volumes into the hpss system, making them known to the pvl server. To make them known to the core server so they can be used by hpss, storage resources must then be created for the volumes via the create tape resources window. When the max number of drives ...

  • Page 227

    A successful import and goes on to the next volume in the list. This makes it easy to restart a partially completed import (after fixing the cause of the error which terminated the first request) by clicking the import button again. There is no need to remove from the list the volumes which were imp...

  • Page 228

    “ab5329” “ab7329” the filling will not occur and an error will be displayed if the specified values would generate an invalid volume label (e.G., one greater than zzz999). To specify a list of volumes from a file, create a file containing the name of each volume to be imported on a separate line. Vo...

  • Page 229

    The end of the volume list. If fill count is greater than 1, multiple labels are generated using the entered label as a starting point. Maximum volumes allowed . The maximum number of volume labels that will fit in the volume list. The value is 10,000 and is set by ssm. This field is non-editable. T...

  • Page 230

    An ansi (non-unitree) or hpss label with a correct volume id (the volume id on the label is as expected by hpss) tape imported label written, tape imported tape imported an ansi or hpss label with an incorrect volume id (the volume id on the label is different from the volume id expected by hpss) ta...

  • Page 231

    This window allows the user to import disk volumes into the hpss system, making them known to the pvl and pvr servers. To make them known to the core server so they can be used, they must be created via the core server's create disk resources window. The ssm system manager processes the volumes one ...

  • Page 232

    List. To automatically generate a list of volume names in the volume list, set the fill count to the desired number of volumes. Set the fill increment to the number by which each automatically generated label should differ from the previous one. Then type the starting volume name into the volume lab...

  • Page 233

    Based on the import type, the import request will be processed depending on how the media is currently labeled. See section 8.1.1.4: selecting import type for disk volumes on page 234 for more information on selecting the appropriate import type. File containing volume list . The name of an external...

  • Page 234

    Again. You may dismiss the window before completion; however, completion messages will be displayed in a pop-up window. At this point you can begin entering data for another import, or you can dismiss the window. 8.1.1.4. Selecting import type for disk volumes the following table lists information f...

  • Page 235

    Are copied to the volumes' metadata and become a permanent part of the definition of the volumes. The core server creates the necessary metadata structures for each of the new virtual volumes. Each new virtual volume is immediately available for use. Note that new tape resources are not assigned to ...

  • Page 236

    List. To automatically generate a list of volume names in the volume list, set the fill count to the desired number of volumes. Set the fill increment to the number by which each automatically generated label should differ from the previous one. Then type the starting volume name into the volume lab...

  • Page 237

    Server. Vvs to create . The number of virtual volumes to be created. This value determines the number of rows in the volume list table at the bottom of the window. Optional or informational fields pvs in each vv . The stripe width of the selected storage class. This field may not be changed. Pv size...

  • Page 238

    This window is used to create disk storage resources, disk virtual volumes, in a core server. The disks must first be imported to the appropriate pvl. The names of the volumes may be entered into the window one at a time, or a list of volume names may be automatically generated from a single entry. ...

  • Page 239

    "aa0080" "aa0090" "aa0100" "aa0110" "aa0120" when an addition produces overflow in a column, numerical columns are carried over properly to alphabetic columns and vice versa. Example: fill count = 6 fill increment = 2000 volume label= "aa7329" labels automatically inserted into volume list table: "a...

  • Page 240

    Available. The default value is 1. If the list fills before the fill count is exhausted, filling stops and a message is displayed. Fill increment . This field determines how each new volume label is generated. The fill increment is the number by which each automatically generated label will differ f...

  • Page 241

    Reuse the volumes, create new storage resources on them using the create resources window. To remove them entirely from the hpss system, export them from the pvl using the export volumes window. 8.2.1.1. Rules for deleting resources volumes on which resources are to be deleted must be empty. There m...

  • Page 242

    Volume list at the bottom of the window. Any of the three entry methods may be repeated multiple times on the same window to add additional volumes to the list. All three entry methods or any combination of them may be used in succession on the same window. To add a single volume name to the volume ...

  • Page 243

    The file in the file containing volume list field. The volume names from the file will be added to the volume list. Field descriptions file containing volume list. The name of an external file containing a list of volume labels to be added to the end of the volume list. Fill count. The number of vol...

  • Page 244

    8.2.2.1. Rules for exporting volumes tape cartridges may be physically exported from any managing robotic library. To export a tape cartridge from hpss, the administrator must be familiar with the operation of the pvr from which the cartridge will be removed because the physical cartridge ejection p...

  • Page 245

    This window allows you to export tape and disk volumes from the hpss system. Exporting a volume is equivalent to telling hpss that the volume no longer exists. Before volumes can be exported, the core server storage resources that describe the volumes must be deleted using the procedure described in...

  • Page 246

    Fill count = 6 fill increment = 10 volume label = "aa0070" labels automatically inserted into volume list: "aa0070" "aa0080" "aa0090" "aa0100" "aa0110" "aa0120" when an addition produces overflow in a column, numerical columns are carried over properly to alphabetic columns and vice versa. Example: ...

  • Page 247

    Field descriptions eject tapes after exporting . If this checkbox is selected, the exported tape volumes will also be ejected from the pvr. File containing volume list. The name of an external file containing a list of volume labels to be added to the end of the volume list. Fill count . The number ...

  • Page 248

    Available to the storage class. The migration and purge policies may need to be modified to free up more space or to free up the space more frequently. In addition, the total storage space for the storage class may need to be reviewed to determine whether it is sufficient to accommodate the actual u...

  • Page 249

    To select a row, click on it with the mouse; the selection will be highlighted. Note that when you select a row, you are selecting a storage class within a particular storage subsystem. See also the related window configured storage classes , described in section 6.1.1 on page 157. The configured st...

  • Page 250

    Appropriate when the mps is recycled. Possible values are: • waiting - migration is not taking place at this time. The start of the next migration is waiting until criteria specified in the migration policy are met. • running - a migration is in progress. • none - the storage class, as configured, i...

  • Page 251

    • suspend - if the purge state is waiting or running, this puts the purge into the suspended state. • resume - if the purge state is suspended, this returns it to waiting and allows mps to again begin scheduling purge runs. • reread policy - tells the mps to refresh its purge policy information by r...

  • Page 252

    This window allows you to view and update the information associated with an active disk storage class. It reports the storage space data as well as any exceeded thresholds. The window also reports detailed information on the migration and purge status. In addition, the window allows the ssm user to...

  • Page 253

    Field descriptions storage class name . The name assigned to this storage class. Storage class id . The numeric id of this storage class. Storage class type . The class of media assigned to this storage class (tape or disk). Subsystem name . The name of the storage subsystem for which the storage cl...

  • Page 254

    Alarm to ssm. Critical threshold . For disk storage classes, this value is a percentage of total storage space. When the used space in this storage class exceeds this percent of the total space, the mps will send a critical alarm to ssm. Note that the disk threshold reaches the warning level when th...

  • Page 255

    Suspended state. • resume - if the migration state is suspended, this returns it to waiting and allows mps to again begin scheduling migration runs. • reread policy - tells the mps to refresh its migration policy information by rereading the policy. Pending operations . When the mps cannot respond i...

  • Page 256

    8.3.3. Mps tape storage class information this window allows you to view and update the information associated with an active tape storage class. It reports the storage space data as well as any exceeded thresholds. The window also reports detailed information on the migration status. In addition, t...

  • Page 257

    Field descriptions storage class name . The name assigned to this storage class. Storage class id . The numeric id of this storage class. Storage class type . The class of media assigned to this storage class (tape or disk). Subsystem name . The name of the storage subsystem for which the storage cl...

  • Page 258

    Administrator. Start time . The date and time when the most recent migration run started. It may still be running. End time . The date and time when the last migration run completed. Total units processed . The amount of space in the storage class which has been migrated during the current or or mos...

  • Page 259

    Core server metadata that describes the volumes. See section 8.4.3, repacking and reclaiming volumes . 8.4.1. Forcing migration the migration purge server runs migration periodically in the time interval specified in the migration policy. However, between these automatic migration runs, an administr...

  • Page 260

    Repack selects tape volumes in one of two ways. The administrator can provide a list of tape volumes to repack, or repack can select volumes based on a number of selection criteria. If repack is provided with a list of tape volumes to process, those volumes must be in rw, ro, eom or empty condition....

  • Page 261

    This window provides an interface for repacking tape virtual volumes. Field descriptions storage class name. The name of the storage class that will be repacked. Storage class id. The id of the storage class that will be repacked. Subsystem id. The id of the subsystem which contains the volumes to b...

  • Page 262

    Which a large percentage of the files have been deleted, may be selected by repack . If vv space is 100, the comparison is not performed and selection of tape volumes to repack is made using the remaining criteria. Repack options . • select only retired vvs . If selected, repack selects only retired...

  • Page 263

    Buttons reclaim. Press this button to start the reclaim utility program on the indicated storage class. Tape volumes that are described as empty, and are not retired, will be reclaimed. Status messages are displayed on the status bar at the bottom of the window at the start and end of the reclaim. S...

  • Page 264

    Imported. Cs volume . Once you have filled in both fields, clicking on this button will open the core server disk volume or core server tape volume window for the specified volume. This metadata is created when the disk/tape storage resources are successfully created. 8.5.2. Pvl volume information w...

  • Page 265

    • unallocated. The volume has been successfully imported into the hpss system and labeled. However, no storage resources have been created on it and it has not been allocated to a core server. It is therefore not available to the system for i/o. • allocated - on shelf. Storage resources have been cr...

  • Page 266

    Hpss management guide november 2009 release 7.3 (revision 1.0) 266.

  • Page 267

    This window allows you to view and update the information associated with an hpss tape cartridge. Note that the location type fields are represented differently for certain types of robots, for which port , drive and slot ( unit , panel , row , and column ) may each be displayed as 0. If it is neces...

  • Page 268

    Note that this can also mean that side 0 of the cartridge is mounted. Pvr server . The descriptive name of the pvr which manages the cartridge. • cartridge type . The hpss media type corresponding to this cartridge. This controls which the type of drive in which the cartridge can be mounted. Manufac...

  • Page 269

    • port - the location is a port number. [this option is currently not used.] • drive - the location is a drive id number. • slot - the location is a slot specification. The following fields are filled in (non-zero) based on the location type and whether or not the pvr has the information: port . The...

  • Page 270

    This window displays information about a disk volume as represented by the core server. Field descriptions name. The ascii name of the first physical volume that is a part of the disk virtual volume. The entire virtual volume can be referred to by this name. Vv condition . This is the administrative...

  • Page 271

    Volume. • in down condition, the volume cannot be read, written or mounted. This condition can be used to make a disk unavailable to the system. Change the vv condition of a disk virtual volume by selecting the desired condition from the drop down menu and then pressing the update button. Changes pe...

  • Page 272

    Less commonly used data tab actual length . The length of disk virtual volume in bytes. This length includes all of the space set aside for system use. See usable length . Pv size. The length in bytes of a physical volume in this disk virtual volume. All physical volumes in a disk virtual volume mus...

  • Page 273

    Physical volumes this is a table of physical volume attributes for the physical volumes that make up this disk virtual volume. Vol name. The ascii name of the physical volume. Type. The media type. Dev id. The id of the device the physical volume is mounted on. Mvr ip addr. The ip address of the mov...

  • Page 274

    This window displays information about a tape volume as represented by the core server. Field descriptions name. The ascii name of the first physical volume that is a part of the tape virtual volume. The entire virtual volume can be referred to by this name. Vv condition . This is the administrative...

  • Page 275

    • in rwc condition, the volume can be read and written. This is the normal operational state. • in rw condition, the volume can be read and written, but new tape storage segments may not be created on the volume. • in ro condition, the volume can be read but not written. New storage segments cannot ...

  • Page 276

    Its data. When data is being written to a volume, it will be shown in allocated state. When the end of media marker is reached during a tape write operation, the volume will enter eom state. If the number of segments on the volume drops to zero after reaching eom, the state will change to empty. The...

  • Page 277

    Of time may pass between updates. Next write address. The next address that will be written on the tape volume, expressed as an hpss relative stripe address. Less commonly used data tab file family. The family to which the volume is assigned, if any. If it is not assigned to a family, it is assigned...

  • Page 278

    Vv block size. The virtual volume block size. This is the number of bytes written from a data stream to an element of the striped volume before the stream switches to the next element of the stripe. Pv block size . The size, in bytes, of the media data block. Max blocks between tape marks . The maxi...

  • Page 279

    Vv condition controls the availability of the volume for the following actions: • creation of new storage segments • reading of existing storage segments • writing of existing storage segments • mounting of tape media tape volumes have six possible settings for vv condition: • rwc - read, write, cre...

  • Page 280

    Can be read, but not written. Unlike ro condition, tapes in eom condition can only be changed to down. Eom volumes cannot enter either rwc or ro condition. Volumes in down condition cannot be read, written, created on or mounted. This setting effectively removes the volume from the system while main...

  • Page 281

    This window allows you to change which pvr owns a set of cartridges. Before initiating the request from the ssm window, the cartridges must already be physically placed into a tape library managed by the new pvr. The list of the volumes to be moved may be constructed in any of three ways. Each volum...

  • Page 282

    Fill increment = 10 volume label = "aa0070" labels automatically inserted into volume list: "aa0070" "aa0080" "aa0090" "aa0100" "aa0110" "aa0120" when an addition produces overflow in a column, numerical columns are carried over properly to alphabetic columns and vice versa. Example: fill count = 6 ...

  • Page 283

    Fill count . The number of cartridge labels to be added to the end of the list when the volume label field is next modified. This number may be one or greater. If the list fills up before the fill count is exhausted, filling stops, and a message box is displayed (see maximum volumes allowed below). ...

  • Page 284

    8.6.1. Pvl job queue window this window shows all outstanding jobs in the pvl. From this window, the user can issue a request to view more information for a particular pvl job or to cancel it. Each pvl job represents a volume mount (or series of mounts for a striped disk or tape volume). Field descr...

  • Page 285

    • relabel - a cartridge being relabeled. • sync mount - a synchronous mount. • tape check-in - a cartridge being added to the library. • tape check-out - a cartridge being removed from the library to be placed on the shelf or vault. Status . The current status of the job. Possible values are: • unco...

  • Page 286

    8.6.2. Pvl request information window this window is displayed when the job info button is pressed on the pvl job queue window. It allows you to view the information associated with a pvl job/request. Field descriptions job id . The unique number assigned to the job being viewed. Request type . The ...

  • Page 287

    • aborting - the job is being aborted. • cartridge wait - the job is waiting for another job to release a cartridge that it needs. • completed - the job is completed. Once a job is completed, it no longer exists in the pvl job queue, and this window will no longer receive any updates. • deferred dis...

  • Page 288

    • tape check-in • tape check-out • uncommitted • unload pending drive pool id. If non-zero the drive pool id will restrict this drive's scheduling to tape requests specifying this value. This field is not applicable for disks. Drive type . The hpss drive type assigned to the drive. Mover. The name o...

  • Page 289

    Will begin logging alarms indicating that the appropriate tape has yet to be checked-in. The frequency of these alarms is controlled by the shelf tape check-in alarm field of the pvr-specific configuration window. The requests are displayed in chronological order. Each time such a request is receive...

  • Page 290

    The tape mount requests window displays tapes which need to be mounted in a drive. All hpss tape mount requests, including both robotic and operator tape mounts, will be displayed in the window. For operator pvrs, such mount requests mean that a tape must be mounted by hand. When mount requests for ...

  • Page 291

    Clear list . Clears the list of mount requests. Note that this does not cancel any mount requests, but just removes them from the list. Pending mounts will reappear in the window as the pvr periodically retries the mounts. This can be useful for removing stale mount requests that, for some reason, n...

  • Page 292

    Refresh their caches. 3. Use the retire utility program to retire the old technology volumes. R etire can accept a list of pvs to retire, or can retire all volumes in a storage class that are in a specified condition. Dump_sspvs can be used to create a list of pvs for retire to process. Creating a l...

  • Page 293

    Hpss management guide november 2009 release 7.3 (revision 1.0) 293.

  • Page 294

    Chapter 9. Logging and status 9.1. Logging overview the purpose of logging is to record events of interest that occur in hpss in the sequence they occur to support diagnostic research. Hpss provides eight log message types: • alarm • event • status • debug • request • security • accounting • trace t...

  • Page 295

    5. The ssm alarms and events window (section 5.2.2.3: on page 151) a standard configuration for logging services is usually set by the administrator during the hpss system configuration. Specialized configurations can be set up and used to temporarily (or permanently) provide more or less logging fo...

  • Page 296

    • if no server-specific logging policy is configured for a server and no default logging policy is configured, only alarm and event messages will be logged. 9.2.2. Logging policies window this window is used to manage all the log policies in the hpss system. To create a new log policy, click on the ...

  • Page 297

    Delete. Deletes the selected log policy(s). 9.2.2.1. Logging policy configuration window the logging policy window is used to manage a log policy. When creating a new log policy, the descriptive name field will be blank, and a set of default options will be selected. If there is a default logging po...

  • Page 298

    Recommended that this always be selected. • event. An informational message (e.G., subsystem initializing, subsystem terminating) about a significant occurrence in the system that is usually not an error. It is recommended that this always be selected. • request. A message which reports the beginnin...

  • Page 299

    A server’s log policy can be modified to control the volume of messages to the chosen logging destinations. Typically, during normal operations, the level of logging may be decreased to only alarm, event, and security to reduce overhead. However, while tracking an hpss problem, it may be desirable t...

  • Page 300

    9.3.2. Viewing the central log (delogging) hpss provides the ability to retrieve and examine hpss log records as a means of analyzing the activity and behavior of hpss. The retrieval and log record conversion process is referred to as “delogging.” delogging is the process of retrieving specific reco...

  • Page 301

    The information is acquired from the hpss log daemon. The log files information window provides information about the hpss central log files. Log file information includes the state of each log file, the current size of each log file in bytes, the time at which each log file was marked in use, the t...

  • Page 302

    9.5.1. Configuring local logging options the “log messages to:” field in the logging client specific configuration window (section 5.1.4: log client specific configuration on page 100) can be modified to control the destinations for the messages logged by the hpss servers running in a node. This par...

  • Page 303

    This window displays a number of the most recent alarm and event messages which have been received by ssm. It also allows you to view individual messages in greater detail by selecting the message and pressing the alarm/event info button to bring up the alarm/event information window. Field descript...

  • Page 304

    This window displays all the details of the alarm or event selected from the alarms and events window. Field descriptions id . A sequence number assigned to the log message by ssm. This id is not used outside of ssm. Log type . General class of the message. May be either “alarm” or “event”. Event ti...

  • Page 305

    • critical • indeterminate • cleared these may be accompanied by a color status indicator: • (red) - critical or major alarm • (yellow) - minor or warning alarm • none – events and other alarm types error code . The error code associated with the problem underlying the message. Msgid . The 8-charact...

  • Page 306

    Environment variable hpss_ssm_alarms with the desired name of the cache file. The default for hpss_ssm_alarms is defined in hpss_env_defs.H as null. Ssm will revert to the internal memory cache if it cannot access the specified cache file for any reason. The site may set the hpss_ssm_alarms environm...

  • Page 307

    Could be used, for example, to reduce the size of each data request on a slow network. The internal cached alarm and event list is displayed by the hpssadm program by means of its "alarm list" command. This command has a "-c" option to specify how many of the most recent log messages in the internal...

  • Page 308

    Chapter 10. Filesets and junctions a fileset is a logical collection of files that can be managed as a single administrative unit, or more simply, a disjoint directory tree. A fileset has two identifiers: a human readable name, and a numeric fileset id. Both identifiers are unique to a given hpss re...

  • Page 309

    File family. The name of the file family to which the fileset is assigned. If this field contains "not in a family", the fileset has not been assigned to a family. Fileset id . The id number which identifies the fileset. A fileset id is displayed as two double- comma-separated unsigned integer numbe...

  • Page 310

    10.2. Creating an hpss fileset this section provides information on how to create hpss filesets. Only the hpss root user and ssm principal are allowed to create filesets. In order to successfully perform fileset administration, the db2 helper program must be bound. See the hpss installation guide, s...

  • Page 311

    Field descriptions fileset name . The name to be assigned to the fileset. This name must be unique to the realm in which hpss resides. Fileset state . The state of the fileset. If read is on, the fileset will be available for reading. If write is on, the fileset will be available for writing. File f...

  • Page 312

    10.3. Managing existing filesets this section describes how to look up information on, modify, or delete filesets. 10.3.1. Core server fileset information window this window allows an administrator to view, update and/or delete the core server information associated with a fileset. This information ...

  • Page 313

    The data or metadata. Changing the state to destroyed will prevent both reading and writing. Field descriptions fileset id . The id number which identifies the fileset. A fileset id is displayed as two double- comma-separated unsigned integer numbers. A new fileset id can be entered as two double- c...

  • Page 314

    Uid . The user id identifying the user owning the root node of the fileset. Gid . The group id identifying the principal group owning the root node of the fileset. Permissions . The unix-style permissions assigned to the root node of the fileset. There are nine checkboxes arranged in a matrix with t...

  • Page 315

    10.5. Creating a junction only the hpss root user and ssm principal are allowed to create junctions. A junction is a name space object that points to a fileset and is similar to a persistent unix mount point. The fileset pointed to may reside in another subsystem. Junctions can be created using ssm ...

  • Page 316

    10.6. Deleting a junction junctions can be deleted using ssm or by using the utility routine deljunction . For more information, refer to the deljunction man page. To delete a junction using ssm, select the junction(s) to be deleted from the filesets & junctions list and press the delete junction bu...

  • Page 317

    Chapter 11. Files, directories and objects by soid this chapter describes two ways to display basic information about files and directories stored in hpss. Starting with a fully qualified path name, you can look up a file or directory in the system and display information about it. In other cases, y...

  • Page 318

    11.1.1. File/directory information window this window shows details about a file or directory. If a file is displayed in the file/directory information window, a button labeled show bitfile id will appear at the bottom of the window. Pressing this button will cause the storage object id window to ap...

  • Page 319

    Object type . The type of object being displayed, either file or directory. Class of service . The name of the class of service in which the file is stored. If the displayed object is a directory the value of this field will be none. File family . The name of the file family to which the file has be...

  • Page 320

    Extended acl . If any acl entry other than the default acl entries exist, then the file or directory is said to contain extended acls. There are three type of acls that could have extended acls: • object acl – hpss name space object acl • ic acl – hpss name space initial container acl • io acl – hps...

  • Page 321

    Comment x x compositeperms x x cosid x datalength x x entrycount x extendedacls x x familyid x filesethandle filesetid x x filesetrootid x x filesetstateflags x x filesettype x x gid x x groupperms x x linkcount x modeperms x x name x x opencount x optionflags x otherperms x x readcount x subsystemi...

  • Page 322

    11.2. Objects by soid window to display this window, select monitor from the health and status window, then select lookup hpss objects, and then“ objects by soid . This window allows you to open an information window for an hpss object which you specify by the object's hpss storage object id (soid)....

  • Page 323

    Chapter 12. Tape aggregation this chapter discusses the following operations: • overview of tape aggregation • tape aggregation performance considerations • configuring tape aggregation 12.1. Overview of tape aggregation tape aggregation reduces file processing time when migrating relatively small f...

  • Page 324

    Migration policy screen. Edit any other tape aggregation related fields on that screen as needed. If mps is running, you must also tell it to reread the disk migration policy. Hpss management guide november 2009 release 7.3 (revision 1.0) 324.

  • Page 325

    Chapter 13. User accounts and accounting 13.1. Managing hpss users after the hpss system is up and running, the administrator must create the necessary accounts for the hpss users. For a new hpss user, a kerberos, ldap, or unix id (depending on authentication type configured) and an ftp id must exis...

  • Page 326

    [ added unix user ] [ kadmin_princ unset; using kadmin.Local for kerberos ops ] [ adding kerberos principal ] [ added kerberos principal ] hpss/ldap home directory: / primary group id#: 210 [ adding ftp user ] [ ftp user added ] [ adding ldap principal ] [ added ldap principal ] [ adding ssm user ] ...

  • Page 327

    13.1.1.3. Add a kerberos user id the hpssuser utility invokes the kadmin utility to create the krb principal and account. This can be done using both keytab and password by specifying the -krbkeytab option. Invoke the hpssuser utility as follows to add a krb user id using a keytab and password: hpss...

  • Page 328

    # hpssuser -add user1 -ftp -nohome user id#: 300 enter password for user1: ****** re-enter password to verify: ****** full name: test user hpss/ldap home directory: /home/user1 login shell: /bin/ksh primary group id#: 210 [ adding ftp user ] [ ftp user added ] if the -nohome option is not specified ...

  • Page 329

    [ ssm user deleted ] [ deleting ldap principal ] [ deleted ldap principal ] [ deleting ftp user ] [ ftp user deleted ] [ deleting kerberos principal ] [ kadmin_princ unset; using kadmin.Local for kerberos ops ] [ deleted kerberos principal ] [ deleting unix user ] 13.1.3. Listing hpss users the hpss...

  • Page 330

    # hpssuser -ssmclientpkg /tmp/ssmclientpkg.Tar [ packaging ssm client ] [ creating /tmp/ssmclientpkg.Tar ] ssm.Conf krb5.Conf hpssgui.Pl hpssgui.Vbs hpss.Jar [ packaged ssm client in /tmp/ssmclientpkg.Tar ] 13.2. Accounting hpss maintains accounting information on the usage of the system whether the...

  • Page 331

    An accounting policy is required whether the site actually charges users for hpss usage or not. Field descriptions accounting style. The style of accounting that is used by the entire hpss system. Valid values are site or unix. The default value is unix. Under unix style accounting, resource usage i...

  • Page 332

    Manipulation operations. Account inheritance. A flag that indicates whether or not newly created files and directories should automatically inherit the account index used by their parent directory. The default value is off. It is only used if account validation has been enabled and site- style accou...

  • Page 333

    This window allows an administrator to view the accounting status and start accounting. Field descriptions subsystem. The name of the storage subsystem containing this accounting status data. Run status. Current status of accounting run. Possible values are: • never run • running • failed • complete...

  • Page 334

    The first type, denoted by a zero (0) in the fourth column, gives the following summary information about the storage used by a particular hpss account index (acctid) in a particular class of service (cos): • the total number of file accesses (#accesses) to files owned by the account index in the cl...

  • Page 335

    In the above example, line 2 shows that a user using account 634 made a total of 89 accesses to cos 1 and has 125 files stored in cos 1 which together total 4168147 storage units. The storage units reported by the report utility may be configured in the accounting policy to represent bytes, kilobyte...

  • Page 336

    ● the site will need to create a local site account map to maintain a list of the account ids for which each user is authorized. This is a locally designed and written table, not supported by hpss. See section 13.2.3.1: site defined accounting configuration files and procedures on page 336 for sugge...

  • Page 337

    13.2.3.1.2. Site defined account apportionment table in unix-style accounting, the uid (as account index) maps only to a specific user. Some sites may wish to apportion different percentages of the charges for a single uid among different project charge codes, but without using site style accounting...

  • Page 338

    13.2.3.2. Accounting intervals and charges the time between accounting runs and the charging policy for space usage should be developed after consulting the accounting requirements. The following are some guidelines to consider: • accounting should be run at a regular intervals, such as once per mon...

  • Page 339

    Chapter 14. User interfaces this chapter configuration information for the user interfaces provided with hpss for transferring files: • client application programming interface (api) • parallel file transfer protocol (ftp) or pftp • hpss virtual file system (vfs) interface 14.1. Client api configura...

  • Page 340

    Write operations between cache invalidates. The default value is 20. Hpss_api_hostname specifies the hostname to be used for tcp/ip listen ports created by the client api. The default value is hpss_host. This value can have a significant impact on data transfer performance for data transfers that ar...

  • Page 341

    Etc/hpss.Keytab. Hpss_unix_keytab_file specifies the name of the file containing the security keys necessary for successfully initializing the client api for unix authentication. The default is auth_keytab:/var/hpss/etc/hpss.Unix.Keytab. 14.2. Ftp/pftp daemon configuration the pftp_client binary is ...

  • Page 342

    Non-hpss parallel ftp daemon (dis 2 com pftp daemon). This file should be customized as needed. Refer to the hpss.Conf man page or the hpss installation guide , appendix d for details. Note: it may be necessary for a site to merge older copies of the hpss.Conf file with the template if modifications...

  • Page 343

    # hpss file systems rather than the local file system of the # host executing the ftp daemon. This is highly recommended! Banner # control for logging (sent to syslog()). Log [ commands ] [ anonymous ] [ { inbound } ] [ transfers ] [ guest ] [ { outbound } ] [ real ] [debug] # set the maximum number...

  • Page 344

    • message/readme/banner/shutdown (message lines) are text files, with the following keywords (all one character in length) recognized, identified by a preceding % : table 5. Banner keywords keyword description m class limit t current time on ftp daemon c current working directory r remote hostname l...

  • Page 345

    Days of the week. Step 4. Creating ftp users in order for an hpss user to use ftp, a unix and/or kerberos userid and password must be created. Refer to section 3.3.2.1: the hpssuser utility on page 35 for information on how to use the hpssuser utility to create the userid and password and set up the...

  • Page 346

    Character device (/dev/hpssfs0) that is used to communicate with the hpss vfs daemon (hpssfsd). All posix i/o system calls like open, read, write, close, etc. Are first handled by the vfs abstraction layer after which they are passed down to appropriate functions in the kernel module. The kernel mod...

  • Page 347

    % make build_root=/tmp/vfs_client build-clnt build-fs create a tar file that can be used to build the client code on the vfs machine: % cd /tmp/vfs_client % tar -cvf ../vfs_client.Tar * copy (i.E. Scp) the new tar file to the client machine. On the client machine, untar and build the client tree. % ...

  • Page 348

    % rmmod hpssfs # this ensures that there isn't a pre-existing module loaded % modprobe hpssfs % make config % makedev hpssfs % /sbin/chkconfig –add hpssfs build and install the application daemon by following the instructions given by the build-help target of the makefile in the directory above this...

  • Page 349

    % mkdir /var/hpss/cred % mkdir /var/hpss/tmp on the core server machine, use mkhpss to create the client config bundle: % mkhpss select "create config bundle" to create a client config bundle that contains config. Files from the core server machine: [ adding hpss local passwd/group/shadow files to b...

  • Page 350

    O /opt/hpss/lib/libhpssunixauth.So o /opt/hpss/lib/libhpssldapauthz.So • if the code is installed in a non-standard location (not /opt/hpss), update the paths in auth.Conf and authz.Conf to use the correct location. If using kerberos authentication, modify the /etc/krb5.Conf file on the client machi...

  • Page 351

    14.4. Mounting vfs filesystems an hpss fileset or directory is made available for user access by mounting it using the mount(8) command. The mount command accepts the mount input options directly from the command line or from the corresponding entry defined in the /etc/fstab file. By defining the mo...

  • Page 352

    None /dev/shm tmpfs defaults 0 0 none /proc proc defaults 0 0 none /sys sysfs defaults 0 0 /dev/volgroup00/logvol01 swap swap defaults 0 0 jupiter:/auto/home /home nfs defaults 0 0 local:/ /tmnt hpssfs noauto,cos=2 0 0 local:/home/user1 /home/hpss/user1 hpssfs noauto,maxsegsz 0 0 remote- site:/home/...

  • Page 353

    Mount option mount description acdirtimeo seconds to cache directory attributes. The shorter this value, the more frequently communication to hpss is required. Synchronization between separate vfs nodes or mount points may benefit for a shorter cache period with the expense of longer latencies. Flus...

  • Page 354

    Mount option mount description maxiowidth maxiowidth princ override for principal auth override for authorization type key override keytab type keytab override for keytab file stage / nostage default is stage. This overrides the cos setting. Application can override this by specifying o_nonblock on ...

  • Page 355

    Where can be obtained using the mount utility. The trace level can be changed while the hpssfsd is running. # echo 5 > /proc/fs/hpssfs//trace • show general information and statistics for a mounted file system: # cat /proc/fs/hpssfs//info • list files currently opened for a mounted file system: # ca...

  • Page 356

    Chapter 15. Backup and recovery this chapter discusses the following operations: • backup and recover hpss metadata • backup hpss environment • recover hpss user data • handling db2 space shortage 15.1. Hpss metadata backup and recovery each hpss site is responsible for implementing and executing th...

  • Page 357

    Made to the database. These logs allow db2 to recover all changes made to the database since the time of the last database backup, forward to the time when db2 was stopped by a crash, hardware failure, power outage or whatever. It is vital that the db2 log files be hosted on a highly reliable disk s...

  • Page 358

    “full”, “incremental”, or “delta”. Full backups record the full contents of the database at a point in time. Incremental backups record all the changed data since the last full backup. Delta backups record only the data that has changed since the most recent backup (of any type). Obviously, a full b...

  • Page 359

    15.1.3. Overview of the db2 recovery process upon discovering a damaged db2 container, you must first determine the level of hardware recovery and problem determination available to you. For example, whether or not you were utilizing some level of raid and can recover a failed disk from a good disk....

  • Page 360

    • {hpss secondary metadata backup path}/subsys1 • / • others site-specific filesystems 15.2.2. Operating system backup it is also necessary to perform appropriate os backups on a regular basis. However, these activities are outside the scope of this document. Refer to the relevant os documentation f...

  • Page 361

    Prepare for and to perform the recovery process: 1. Determine the name of the potentially damaged volume. Attempts to read the damaged volume will result in mover alarm messages being issued to ssm. The alarm messages will contain the name of the physical volume for which the error occurred. Record ...

  • Page 362

    Can be used to clean up storage resources and report what data has been lost. The recover utility has several options. These options are used depending on the severity of the damage. Refer to the following sections for more information on how and when to use these options. The recover utility can on...

  • Page 363

    Storage level = 1 on vv = vol00100 path ( fileset24: /home/bill/file2) ========= trying to recover bitfile ========= 0786ab2c-156b-1047-8c81-02608c2f971f 00336b52 4631 10cf 00 00 00 02 storage level = 1 on vv = vol00100 path ( fileset24: /home/bill/file3) ========= trying to recover bitfile ========...

  • Page 364

    Storage level = 1 on vv = vol00100 path ( fileset24: /home/bill/file2) lost segments from this storage level offset = 0 , length = 32768 offset = 32768, length = 32768 offset = 65536, length = 32768 at the end of the recovery, no segments or volumes associated with the damaged segments are purged or...

  • Page 365

    00226b52 4631 10cf 00 00 00 17 at the end of the cleanup, all the virtual and physical volumes associated with the targeted volumes will be deleted. All the physical volumes contained in the virtual volume will be exported from the pvl. The media can then be imported back into hpss to be reused. At ...

  • Page 366

    15.5.1. Dms table spaces capacity for a dms table space is the total size of containers allocated to the table space. When a dms table space reaches capacity (depending on the usage of the table space, 90% is a possible threshold), you should add more space to it. The database manager will automatic...

  • Page 367

    Hpss management guide november 2009 release 7.3 (revision 1.0) 367.

  • Page 368

    Chapter 16. Management tools 16.1. Utility overview hpss provides a variety of command-line utility programs to aid in the use and management of the system. The programs can be grouped into the following major categories: 16.1.1. Fileset and junction management these programs work for hpss filesets....

  • Page 369

    • dumppv_pvr - list the physical volumes managed by a particular pvr. • lshpss - lists various views of the system configuration. • lsrb - lists a list of bitfiles that have not been accessed since a given date/time. • mps_reporter - produces human-readable output from mps summary files. • plu - lis...

  • Page 370

    • hpss_unix_user - manages hpss unix password file. • hpss_managetables – creates, deletes and modifies databases, database tablespaces, tables, views and constraints. This program is normally used by mkhpss . It should be used by hand only by extremely knowledgeable administrators. This program can...

  • Page 371

    Can also be used to reset the number of segments counter in disk and tape storage maps. • showdiskmaps – sends a command to the core server in the selected storage subsystem to dump its in-memory disk space allocation maps, then displays that information on standard output. Hpss management guide nov...

  • Page 372

    Hpss management guide november 2009 release 7.3 (revision 1.0) 372.

  • Page 373

    Appendix a. Glossary of terms and acronyms aci automatic media library client interface acl access control list acsls automated cartridge system library software (science technology corporation) adic advanced digital information corporation accounting the process of tracking system usage per user, p...

  • Page 374

    Bitfile segment an internal metadata structure, not normally visible, used by the core server to map contiguous pieces of a bitfile to underlying storage. Bitfile service portion of the hpss core server that provides a logical abstraction of bitfiles to its clients. Bmux block multiplexer channel by...

  • Page 375

    Dec digital equipment corporation. Delog the process of extraction, formatting, and outputting hpss central log records. Deregistration the process of disabling notification to ssm for a particular attribute change. Descriptive name a human-readable name for an hpss server. Device a physical piece o...

  • Page 376

    Fileset id a 64-bit number that uniquely identifies a fileset. Fileset name a name that uniquely identifies a fileset. File system id a 32-bit number that uniquely identifies an aggregate. Ftp file transfer protocol gatekeeper an hpss server that provides two main services: the ability to schedule t...

  • Page 377

    Ibm international business machines corporation id identifier iec international electrotechnical commission ieee institute of electrical and electronics engineers ietf internet engineering task force imex import/export import an operation in which a cartridge and its associated storage space are mad...

  • Page 378

    Local log an optional circular log maintained by a log client. The central log contains formatted messages from all enabled hpss servers residing on the same node as the log client. Location server an hpss server that is used to help clients locate the appropriate core server and/or other hpss serve...

  • Page 379

    Mount an operation in which a cartridge is either physically or logically made readable and/or writable on a drive. In the case of tape cartridges, a mount operation is a physical operation. In the case of a fixed disk unit, a mount is a logical operation. Mount point a place where a fileset is moun...

  • Page 380

    Physical volume an hpss object managed jointly by the core server and the physical volume library that represents the portion of a virtual volume. A virtual volume may be composed of one or more physical volumes, but a physical volume may contain data from no more than one virtual volume. Physical v...

  • Page 381

    Request a log record message type used to log some action being performed by an hpss server on behalf of a client. Risc reduced instruction set computer/cycles rms removable media service rpc remote procedure call scsi small computer systems interface security a log record message type used to log s...

  • Page 382

    Stage to copy file data from a level in the file’s hierarchy onto the top level in the hierarchy. Start-up an hpss ssm administrative operation that causes a server to begin execution. Status a log record message type used to log processing results. This message type is being used to report status f...

  • Page 383

    System manager the storage system management (ssm) server. It communicates with all other hpss components requiring monitoring or control. It also communicates with the ssm graphical user interface (hpssgui) and command line interface (hpssadm). Tb terabyte (2 40 ) tcp/ip transmission control protoc...

  • Page 384

    Hpss management guide november 2009 release 7.3 (revision 1.0) 384.

  • Page 385

    Appendix b. References 6. 3580 ultrium tape drive setup, operator and service guide ga32-0415-00 7. 3584 ultrascalable tape library planning and operator guide ga32-0408-01 8. 3584 ultrascalable tape library scsi reference wb1108-00 9. Aix performance tuning guide 10. Data storage management (xdsm) ...

  • Page 386

    33. J. Steiner, c. Neuman, and j. Schiller , " kerberos: an authentication service for open network systems ," usenix 1988 winter conference proceedings (1988). 34. R.W. Watson and r.A. Coyne, “ the parallel i/o architecture of the high-performance storage system (hpss), ” from the 1995 ieee mss sym...

  • Page 387

    Appendix c. Developer acknowledgments hpss is a product of a government-industry collaboration. The project approach is based on the premise that no single company, government laboratory, or research organization has the ability to confront all of the system-level issues that must be resolved for si...

  • Page 388

    Hpss management guide november 2009 release 7.3 (revision 1.0) 388.