Log Monitoring known issues and limitations
You might come across a few issues or limitations while you work with PowerVC monitoring feature. However, these issues do not hinder any of the core functionality of the product. These issues will be fixed during future releases.
Listed are a few issues and limitations.
- Compressed log data cannot be collected by Filebeat
- Filebeat does not provide the capability to read compressed (*.gz) log files. It is common for the operating system to compress older logs to make efficient use of file system space. Filebeat cannot read compressed log files. To load them into the OpenSearch you must decompress them manually and temporarily, prior to monitoring installation or reset. After the logs are collected by Filebeat and viewable in OpenSearch Dashboards, you can compress them again.
- OpenSearch Dashboards user interface is not translated
- OpenSearch Dashboards is an open source component that is maintained by a community of users. It does not offer the capability to translate messages and user interface visualizations. As such, any operation of the UI is only available in plain English. If you would like to see this change, you are welcome to contact the OpenSearch open community and request this feature through the proper channels. However, the community may not prioritize such requests. Since most of the functionality that is provided by OpenSearch Dashboards is for the visualization of log entries (which may or might not be already translated, depending on the services and applications that create such entries), graphs over dashboards, and limited to platform administrators, the impact is limited.
- Backup information is stored in the primary / bootstrap host and mounted to other nodes (multi-node)
- OpenSearch backup on a multi-node installation depends on the existence of a shared NFS among
all nodes in the OpenSearch cluster. During PowerVC installation the NTFS share is hosted - by
default - on the primary / bootstrap host (as defined by the PowerVC inventory). As such, users must take
adequate precaution to make sure that the primary / bootstrap host has additional space on the file
system to host the backup data. By default, the backup location is set to
/backupand uses the root file system. To use another location or a separate disk and file system combination please contact PowerVC support team if you need to change this.
- Only two pruning criteria of old data in OpenSearch are supported: age and space
- The data pruning capability is implemented by a
crontabscript. As such, it could be capable of more complex rules than only pruning data older than a determined number of days or pruning data when consumption is over a determined amount of GBs. Users can modify the pruning script to their own needs on a non-supported basis.
- Pruning of old data in OpenSearch depends on the Crontab service and runs hourly
- Pruning of data in the OpenSearch database is performed by a crontab task that runs hourly by
default. Since the criteria is only enforced on an hourly basis, there might be situations when the
data grows in between crontab checks and the criteria might be breached temporarily until the next
crontab check (up to an hour later). In addition, between these runs the file system might be filled
up to more than 90% utilization and in this case the OpenSearch cluster may enter
protectionmode and stops processing data. If this happens the user should (after correcting the filesystem shortage) manually stop and then restart all monitoring stack components to resume data processing.