Kiuwan On-Premises Distributed Troubleshooting Guide – Kiuwan

How can I collect support data?

If you have detected a bug or need to contact Kiuwan support, you can collect valuable information that will help us diagnose your problem by executing:

cd [INSTALLER_DIR]/support
sudo ./collect-support-info.sh

This will gather all relevant diagnostic data and store it under [INSTALLER_DIR]/support/[yyyyMMddhhmmss], where the folder name corresponds to the current timestamp.

Note that the needed time to collect all data depends on the size of your data and could take some minutes.

As the included files may contain sensitive information, you can check if anything that should not leave your organization must be removed from the files before sending them to Kiuwan support.

Once the support files have been collected, compress the folder into a single zip file (please replace the [yyyyMMddhhmmss] with the name of the folder created by the script:

zip -r support-info.zip [yyyyMMddhhmmss]

Where are logs stored?

The following table shows where logs are stored for each Kiuwan On-Premises service:

Service	Location	Folder	File name pattern	Descr.
apacheloadbalancer	[VOLUMES_DIR]/data-local/ApacheLoadBalancer	logs	*.log	Apache logs
wildflykiuwan-f[n]	[VOLUMES_DIR]/data-local/generatedConfig	front/kwf[n]/logs/server	server.log[.yyyy-MM-dd]	Wildfly fronts servers logs
wildflykiuwan-a[n]		analyzers/kwa[n]/logs/server		Wildfly analyzers servers logs
wildflykiuwan-s[n]		schedulers/kws[n]/logs/server		Wildfly schedulers servers logs
wildflykiuwan-f[n]	[VOLUMES_DIR]/data-local/generatedConfig	front/kwf[n]/logs/kiuwan/logs	*.log.[n]	Kiuwan fronts webapp logs
wildflykiuwan-a[n]		analyzers/kwa[n]/logs/kiuwan/logs		Kiuwan analyzers webapp logs
wildflykiuwan-s[n]		schedulers/kws[n]/logs/kiuwan/logs		Kiuwan schedulers webapp logs
wildflykiuwan-a[n]	[VOLUMES_DIR]/data-local/generatedConfig	analyzers/kwa[n]/logs/kiuwan/logs/analysisLogs/kiuwanadmin-2/analysis	*.log	Kiuwan analysis server logs
updaterkiuwan	[VOLUMES_DIR]/data-local/updater	logs/install	[yyyyMMdd_hhmmss]/*.log	Kiuwan installation log
		logs/upgrade/data		Kiuwan data upgrade log
		logs/upgrade/database		Kiuwan database upgrade log
		logs/upgrade/qmm		Kiuwan models manager upgrade log
		logs/updaterManager	updaterManagerTool.log	Kiuwan upgrade tools log
mysqlkiuwan	[VOLUMES_DIR]/data-local/MySql	logs	error.log	MySQL error log

How can I free space on my server?

Before purging any file, make sure that no analyses are running and Kiuwan services are stopped by issuing this command:

cd [INSTALLER_DIR]
sudo ./stop-all.sh

The following files can be purged without affecting the normal operation of Kiuwan On-Premises:

Unneeded stored logs (refer to Where are logs stored? for a complete list of log files).
Local repository data found in [VOLUMES_DIR]/data-local/generatedConfig:
- front/kwf[n]/repos/kiuwanLocalRepository/kiuwanadmin-2
- analyzers/kwa[n]/repos/kiuwanLocalRepository/kiuwanadmin-2
- schedulers/kws[n]/repos/kiuwanLocalRepository/kiuwanadmin-2
Central repository data of already processed analyses found in [VOLUMES_DIR]/data-shared/repos/kiuwanCentralRepository:
- kiuwanadmin-2/reports

Note that when removing the data shown above, historical data relevant for diagnosis and support will not be available. Please consider backing up your data before removing files.

Once the data has been purged, restart Kiuwan services:

cd [INSTALLER_DIR]
sudo ./start-all.sh

My Kiuwan On-Premises local IP address has changed and Redis is not responding anymore

If you have used the default configuration to install your Kiuwan On-Premises (that is, running all the default containers in a single host), it is recommended to keep a fixed IP in the host machine.

Kiuwan On-Premises default installation uses a Redis Cluster made up of six nodes, all running in the same host. Each one of these nodes needs to know the IP where all the other nodes reside (not the internal docker network IP, but the host local network IP). In case of an IP change, it is needed to update the Redis Cluster nodes IP. This is automatically done by the supplied containers, but if enough time passes since the IP change, Kiuwan will refuse to connect to the Redis Cluster after exhausting the defined maximum retries.

In the previous situation, you will see in any wildflykiuwan-f* container logs these traces:

ERROR [io.undertow.request] (default task-14) UT005023: Exception handling request to /saas/zkau: redis.clients.jedis.exceptions.JedisClusterMaxAttemptsException: No more cluster attempts left.

Problem fix for Kiuwan On-Premises version >= 2.8.2004.5

To recover the system, simply restart all containers by running these commands:

cd [INSTALLER_DIR]
sudo ./stop-all.sh
sudo ./start-all.sh

Problem fix for Kiuwan On-Premises version < 2.8.2004.5

If your Kiuwan On-Premises version is older than 2.8.2004.5, you should restart Redis Cluster manually before upgrading to the latest version:

cd [INSTALLER_DIR]
sudo ./stop-all.sh
cd [INSTALLER_DIR]/docker
sudo ./start-infrastructure.sh

Once the infrastructure containers are up, you can proceed with the upgrade (see Kiuwan On-Premises Distributed Upgrade Guide). Once the upgrade is finished, the problem will be gone.

Related to