Archive Subpool Definition panel (Centera)

Archive Subpool Definition panel (Centera)

Archive Subpool Definition panel (Centera)

PE97AS17 ---------------------------------------------------------------------
 Command ===> _________________________________________________________________

 Insert Archive Subpool Definition Last Update:

 Archive Pool Name : QI#O005 Archive Product : B97
 Archive Media Name : CTERA Archive Pool Owner : CUST001

 Subpool is active ===> YES (Y)es,(N)o
 Order for Reload ===> 04 (1-99)
 Archive Retention Period ===> 3660. (1-36500) Days
 Compress Index ===> YES (Y)es,(N)o
 Prefix for Dataset Name ===> BETA97.ARCHC5...................

 CONNECT Parameter Suffix ===> SUFF001.






 Press the ENTER key to insert the definition.
 Press the END key to return to the previous panel.

Fields (Centera)

Field

Description

Subpool is active

Controls whether this archive subpool is used during the archive run (Yes) or not (No)

Order for Reload

Numeric value determining the order in which the archive subpools assigned to this archive pool will be accessed when data is reloaded from the archive

Legal values are 1 (highest) through 99 (lowest).

The default reload order defined in the subpool definitions can be overridden in the archive pool definition. For more information, see "Example: Overriding default order for reload".

Archive Retention Period

Archive retention period in days (max. 36500)

Compress Index

Yes to archive indexes in compressed format (reduces archive space and number of IOs, but increases CPU consumption)

Prefix for DSname

A unique and valid prefix for archive datasets (max. 32 characters)

Important: The prefix is not used in the dataset name when archiving in a Centera archive. However, a valid prefix must still be entered, as this field is defined as a key in the Adabas Audit Data Retrieval database. If you try to enter a prefix that is already in use in another subpool definition (same or different archive pool), this value will be rejected.

CONNECT Parameter Suffix

Establishes a connection between this archive subpool definition and the connection string to be used, which is defined in the B97LSTxx member (LST parameter B97_CONNECT_TO_CENTERA_suffix)

Requirements

The IBM Language Environment (LE) and Centera Runtime library are required for archiving on Centera.

The IBM Language Environment (LE) and the Centera Runtime-library are normally defined in the linklist. If this is not the case in your environment, please supplement the library in the following places:

  • In the Adabas Audit Data Retrieval started task procedure (default name B97STC)
  • In JCL of B97ARC (and/or the respective step of B97DAILY)
  • In JCL of B97DEARC (and/or the respective step of B97DAILY)
  • Under option S.2, page 3 (for JCL generated online)

Optional: Add the dataset or member which contains the Centera environment variables under the DD name BSACFG in the following places:

  • In the Adabas Audit Data Retrieval started task procedure (default name B97STC)
  • In JCL of B97ARC (and/or the respective step of B97DAILY)
  • In JCL of B97DEARC (and/or the respective step of B97DAILY)
  • In the skeletons SE97ARCH and SE97ACLN (and/or the respective step of SE97DALY) in the BETA97.ISPSLIB (for JCL generated online)

Adabas Audit Data Retrieval sets the environment variables specified in the dataset or member when archiving in a Centera archive.

LST parameters

The following LST parameters in the B97LSTxx member control archival in the Centera archive and must be added and/or customized, if necessary:

  • B97_CONNECT_TO_CENTERA_suffix
  • B97_CENTERA_TASKS
  • OBJ_RETRIEVAL_DEVICES
  • B97_ARC_MAX_NUMBER_OF_OBJECTS

Connection string

When connecting to the Centera archive, Adabas Audit Data Retrieval passes the required parameters in the so-called connection string. The maximum length of the connection string is 256 bytes. The Centera API Reference Guide describes the structure of the connection string as well as the parameters which it must contain (IP address, port, userID/password (or PAE-dataset) etc.).

You can store several connection strings in the B97LSTxx member under the keyword B97_CONNECT_TO_CENTERA_suffix. suffix is variable and establishes the connection between a specific archive subpool definition and the corresponding connection string.

Example: If an archive subpool definition of the type CTERA has the value SUBPOOL1 in the CONNECT Parameter Suffix field, then Adabas Audit Data Retrieval uses the connection string which is defined under the keyword B97_CONNECT_TO_CENTERA_SUBPOOL1 when reading or writing archive datasets of this subpool.

Characteristics

The following characteristics apply when archiving to a Centera archive:

  • The calculated archive retention period is converted into seconds, as required by Centera.

    Centera Retention Classes are currently not supported.

  • Adabas Audit Data Retrieval stores datasets with a fixed size of 4 MB in the Centera archive. The names of these datasets begin with AOR.atoken.

    Each Adabas Audit Data Retrieval archive dataset consists of one or more of these 4-MB blocks. The maximum size of an archive dataset can be controlled via the following LST parameter: B97_ARC_MAX_NUMBER_OF_OBJECTS

  • An archive dataset which consists of several blocks can be processed in parallel. The number of parallel activities can be limited with the help of the following LST parameter in the B97LSTxx member:
    • When archiving: B97_CENTERA_TASKS
    • When reloading: OBJ_RETRIEVAL_DEVICES

    Note on storage requirements (Region): An additional 5 MB of storage is required for parallel archival. This means for example, B97ARC additionally requires 20 MB of storage when archiving to a Centera archive if you enter B97_CENTERA_TASKS = 4.

  • When archiving in a subpool of the type CTERA is processed in parallel, the following activities can only be carried out after all 4-MB blocks of an archive dataset have been fully written :
    • Writing the log IRMPRINT for this archive dataset
    • Resetting the archive pending flag of all lists which are stored in this archive dataset

    This concerns all archive subpools of the relevant archive pool, independent of its type (Disk, Odisk, Tape or Ctera), if at least one archive subpool is of the type CTERA.

    This means: If an error causes the archive run to abort, the Archive-Pending flag is not reset for the lists that were already successfully archived in this archive run. The concerned lists are then archived anew in the next archive run. We therefore recommend that you limit the maximum number of 4-MB blocks of the archive datasets with the help of the B97_ARC_MAX_NUMBER_OF_OBJECTS parameter. Note that the maximum number of archive datasets that can be created on a given day is limited to 999.

Error analysis

The LST parameter OBJ_CENTERA_TRACE = YES in the B97LSTxx member causes a detailed log to be written. Use this parameter only during problem resolution after consulting support.