Version 1.0.10.0 release notes (July 2018)

What's new

Web console

• Baseline is now available on some of the personalized homepage widgets. Baseline shows a typical range for comparison. It shows a range of MEAN +/- 1 standard deviation of data samples in the same time period during the last 4 weeks. Note that the baseline is only available when historical data is sufficient for the server to build the baseline. The baseline is available in the following widgets:
  • Database throughput
  • Database contention
  • Database responsiveness
  For more information on the personalized home page, see Personalizing the console home page.
• The type of actions that you can set for a Call Home alert is limited for some alert types. For example, opening a PMR for information events is not available. Any request for changing an alert rule with an unsupported action is returned with an error message.
• Single sign-on is enabled on the web console to access additional applications such as DSX and Spark. You no longer need to log in to these applications when accessing them from IAS web console.

Backup, restore, and disaster recovery

• IBM Integrated Analytics System now supports IBM Spectrum Protect integration for a single rack system, using diskpools with SAN connections to all nodes as well as patch panel set-up, for LAN-Free backup and restore operations. Contact support for more information on installation and configuration.

  You can also read Configuring IBM Spectrum Protect client for LAN-free backups and Configuring IBM Spectrum Protect server for LAN-free backups for 1.0.17 and lower.

• Platform-level backup now supports backing up the configurations for two new components:
  • Web Console Container:

    host_backup --webconsole

  • System history file:

    host_backup --systemhistory

• A new script emc_client is now available for EMC NetWorker customers which allows you to back up your configuration before upgrading, and then restore the configuration when the upgrade is complete. For more information on the command, see DEPRECATED: Backing up and restoring the EMC NetWorker configuration.
• Added a check to make sure that all the EMC NetWorker client NMDA rpms are installed and configured properly before running a backup or restore operation.
• Added a check to make sure that Db2 configuration parameters for LOGARCHMETH2 and LOGARCHOPT2 are set correctly before running backup or restore with EMC NetWorker.
• db_backup now supports running backups to /dev/null as well, to bring the database back to normal state from a backup pending state.
• Added the following backup path validations for db_backup:
  • Allows only /scratch for local/filesystem backups, with a warning that /scratch has only limited space, and a recommendation to set up either SAN or NFS for backups.
  • Checks if the given backup path is shared across all the nodes.
• IAS now supports active-passive disaster recovery through an integrated IBM Spectrum Scale Active File Management solution. The solution provides the ability to perform a file-based snapshot of the active primary server at scheduled intervals for transport to a passive secondary server for failover in the case of a catastrophic event. For more information, see DEPRECATED: File system based disaster recovery solution.

Storage

The storage_setup script has new parameters for moving SAN storage from one system to another, and configuring SAN in multi-HA domains. For more information, see Managing external storage with the storage_setup script.

Platform management

• The apstop command now checks if Db2 backup or restore operations are in progress. If so, the command exits with an error message. If you need to stop the software anyway and interrupt the backup operation, you can re-run the apstop command with the --ignore-backup parameter. There is no equivalent parameter to ignore the restore operation. You must wait for the restore operation to complete before using the apstop command.
• A new event with reason code 152 is now sent whenever Db2 crashes. In the event, the first occurrence data capture (FODC) diagnostic data is delivered.

Security

Single sign-on for the web console

A new system service called apjwtservice is running on the master node of IAS. This service is internally used to support single sign-on for DSX and Spark services that are deployed in IAS.

Components

Resolved issues

For a cumulative readme of all resolved issues since release 1.0.6.0, see Cumulative readme for fixed issues

Known issues

Upgrade may result in 'Down' system state

During an upgrade, the apstart command that starts the appliance software stack may result in a 'Down' system state due to transient resource issues.

Workaround:

1. Confirm the presence of an issue using the following command: grep 'Unmounting file systems' /var/adm/ras/mmfs.log.latest
2. To recover the system, stop the appliance software stack using apstop, followed by apstart to restart.

Upgrade may exit after finishing upgrade of apupgrade component

After completing the upgrade of the apupgrade program itself, the process may exit with an unknown error when attempting to restart.

Workaround:

If the upgrade process exits, and you see the message

Finished the upgrade of apupgrade component. Restarting the upgrade process using the latest apupgrade

immediately followed by

Unhandled error when attempting upgrade

at the end of the upgrade log, restart the upgrade by using the same apupgrade command that you used previously and it will resume.

IAS fails to recover from failover due to a web console container issue

An issue with the web console container may occur which results in IAS getting stuck in state=RECOVERING for a longer time than usual, because the attempt to disable the DSM connection fails.

To verify that this issue has occurred:

1. Find the current master node by running ap node -d.
2. Run the following command:

   grep -rn 'Could not enable dsserver blackout' /var/log/appliance/application/ha/current_master_node/log/ha.log

   If the grep function returns positive results around the time when the recovery started, you have run into this known issue.

Workaround:

Run the following commands:

ap apps disable webconsole
ap apps enable webconsole

The database cannot be accessed after switching to Active Directory

After switching to the Active Directory, user might be unable to connect to database from both the web console and back-end. After logging in to the console, the error

The connection to the database cannot be completed successfully because an incorrect credential is used for the connection. Contact IBM support if the problem persists

is returned.

Workaround:

To access the database after switching to Active Directory, run the apstop and apstart commands. Connection with the database can be completed both from the web console and back-end.

Db2 Warehouse can't start after setting up disaster recovery multiple times

While setting up disaster recovery, if the apdr setup command succeeds for the first time, do not run it again. Otherwise, before running any database operations, you need to:

• Log in as the root user (contact IBM support)
• Check if fileset is set up correctly by running:

  /usr/lpp/mmfs/mmfileset head keystone

• If it points to:

  /opt/ibm/appliance/storage/head/home/db2inst1/db2/keystore.orig

  then run the following command:

  mv /opt/ibm/appliance/storage/head/home/db2inst1/db2/keystore.orig /opt/ibm/appliance/storage/head/home/db2inst1/db2/keystore

Db2 Warehouse is not WRITE resumed after snapshot job failure

In a disaster recovery environment, if the database applications running on the primary node are suspended for a long time, the database might be put in a WRITE-SUSPEND mode and never returned to WRITE-RESUME mode.

Workaround:

Check whether the database is in WRITE-SUSPEND mode by running apdr status -primary. If so, log in as root (contact IBM support) and run the following commands:

docker exec -it dashDB write-resume
rm -f /opt/ibm/appliance/storage/platform/afm-dr/config/.db_in_ws_mode

Upgrade from 1.0.7 and below fails with downgrade error

When attempting to upgrade from versions 1.0.7 and below to versions 1.0.10 and above, upgrade exits with the error

This software does not currently support downgrade.

Workaround:

The apupgrade program must be manually upgraded prior to the actual upgrade.

1. After creating the upgrade directory for this release (for instance /localrepo/1.0.10.0_release) and moving the release bundle there, go to that directory by using cd command and extract the bundle by running

   tar -C EXTRACT -xzvf iias-1.0.10.0_release.tar.gz

2. Install the apupgrade RPM that corresponds to this release by running

   rpm -Uvh EXTRACT/bundle/app_img/apupgrade/apupgrade-1.0.10.0-SNAPSHOT-master-noarch.rpm

3. Proceed with the upgrade process.