AHF Release 23.9

Enhancement to Controlling the Behavior of Oracle Orachk or Oracle Exachk Daemon

AHF 23.9 includes a new command option reset to change the behavior of Oracle Orachk or Oracle Exachk daemon during autostart, autostop, and upgrade.

Command Description

exachk -autostart reset

orachk -autostart reset

ahfctl compliance -autostart reset

Starts and loads the default schedulers.

exachk -autostop unset

orachk -autostop unset

ahfctl compliance -autostop unset

Removes all default unmodified schedulers.

Easier to Manage Audit Dump Logs

AHF Managelogs feature adds ability to manage audit dump logs.

The AHF Managelogs feature purges logs from default locations like the Grid Infrastructure and all Database Automatic Diagnostic Repository (ADR) destinations.

To do this purging Managelogs uses the Database Automatic Diagnostic Repository (ADR), however ADR does not manage audit dump files. As a result, audit dump files can grow and consume too much space.

The Managelogs feature has been expanded to optionally also include management of audit dump files for Grid Infrastructure and Database.

Configure automatic log purging

  • Configure auto purge:
    tfactl set manageLogsAutoPurge=ON
  • Include audit dumps:
    tfactl set managelogs.adump=ON
  • Set the frequency of purging (defaults to 60 mins)
    tfactl set manageLogsAutoPurgeInterval=<n>
  • Configure how old logs must be for them to be purged (default 30 days):
    tfactl set manageLogsAutoPurgePolicyAge=<d|h>

Purge logs on-demand

  • Enable audit dumps:
    tfactl set managelogs.adump=ON
  • Check the usage for audit dump destination
    tfactl managelogs -show usage
  • Check variation for audit dump destination
    tfactl managelogs -show variation
  • Purge audit dump files along with other destinations managed by managelogs:
    tfactl managelogs -purge

Related Topics

Enhancement to ahfctl setupgrade and ahfctl unsetupgrade to Store or Remove autoupdate Configurations

A new option -autoupdate has been added to ahfctl setupgrade and ahfctl unsetupgrade.

  • To store autoupdate configurations, run, for example,
    ahfctl setupgrade -autoupgrade on -swstage /opt/oracle.ahf -frequency 1 -autoupdate on
  • To turn on autoupdate configurations, run:
    ahfctl setupgrade -autoupdate on
  • To turn off autoupdate configurations, run:
    ahfctl setupgrade -autoupdate off
  • To unset autoupdate configurations, run:
    ahfctl unsetupgrade -autoupdate

Faster Creation of Diagnostic Collections with Insights Reports

AHF TFA collections, which include Insights reports are now created faster.

AHF Insights reports can be generated stand-alone using the command ahf analysis create --type insights. Alternatively, a TFA diagnostic collection can be created with an AHF Insights report included by adding the -insight option to the existing -diagcollect command.

This creation of the AHF Insights report often requires analysis of existing zipped diagnostics. Unzipping and processing the collections is CPU intensive and can be slow.

This process of analyzing the diagnostic collection to generate the AHF Insights report has been streamlined and performance improved. Timings will vary based on the type of collection being performed and the systems involved.

An example baseline from testing shows the following improvement, on the time taken to generate the included AHF Insights report:

  • 23.8: tfactl diagcollect -asm -crs- os -tns -insight -last 1h >> 6.8 seconds
  • 23.9: tfactl diagcollect -asm -crs- os -tns -insight -last 1h >> 1 second

Quicker Grid Infrastructure Problem Resolution with CVU Diagnostics

Cluster Verification Utility (CVU) diagnostic files are included in AHF diagnostic collections.

As CVU (Cluster Verification Utility) diagnostic files contain periodic Grid Infrastructure configuration information and critical diagnostic reports they are often required for diagnosis of Grid Infrastructure problems.

AHF now collects all files under the following CVU directories:

  • <GI_BASE>/crsdata/<node>/cvu/diagnostics/cvu_diag_report.txt
  • <GI_BASE>/crsdata/@global/cvu/baseline/cvures/cvusnapshot*.zip

To include the CVU diagnostic files add the -cvu component to the diagcollect command.

For example:
tfactl diagcollect -cvu -last 1h -noclassify

By default, AHF will include CVU in CRS or Database collections for example, both these automatically include CVU diagnostics:

  • tfactl diagcollect -crs -last 1h -noclassify
  • tfactl diagcollect -database orcl -last 1h -noclassify

New Oracle Orachk and Oracle Exachk Best Practice Checks

Release 23.9 includes the following new Oracle Orachk and Oracle Exachk best practice checks.

Oracle Orachk Specific Best Practice Checks

  • Check asmappl.config consistency across nodes for ODA
  • Verify clusterware ADVM volume resources configuration
  • Verify printk logging configuration

Oracle Exachk Specific Best Practice Checks

  • Verify RoCE cabling and switch ports assignment
  • Check file S_CRSCONFIG_<NODE>_ENV.TXT for consistent limit values across all nodes in the cluster
  • Verify DSA authentication is not supported for SSH equivalency

All checks can be explored in more detail via the Health Check Catalogs: