Log Files for Support
Log files, or logs, help the Support team investigate or troubleshoot an issue. This section provides information about the different logs you can create, how to create them, and what information you need to provide to the Support team. If you need assistance, create a case.
One of the most important things you can do when reporting an issue is to provide as much relevant information as possible at the beginning. Some guidelines to consider are:
- What is the problem? Describe the issue as clearly as possible.
- When did the problem occur? Try and narrow it down to a specific session or time of day.
- Provide any other additional information that could assist with solving the issue.
- Provide logs with relevant information that cover your specific use case.
To increase efficiency and reduce the time spent by agents following up with you, Support needs:
- Version of Exasol in use.
- Network information, such as the IP CIDR block, IP subnet, and netmask. .
- IP addresses of all Exasol nodes.
- Number of data nodes.
- Number of storage disks attached per node.
- Backup strategy, and location (local or remote).
- If there are backups, how long does a full backup take?
- Any other important / considerable settings / modifications.
- Database statistics created using
Provide relevant logs to get the following information:
- Cluster information
- Database statistics
- Debug Information
You can retrieve general information about the cluster, such as details about disks, partitions, software versions, and more.
Exasol logs system information in statistical system tables (in the EXA_STATISTICS schema). This information provides insight concerning changes in database behavior. These statistics also provide a good starting point for further system analysis concerning sizing and performance. To know how to download the database statistics, see Download Database Statistics.
Depending on your issue, you need to provide specific debug information. Often this will be requested by the Support agent.
Debug Use Cases
This section provides you with information about different use cases and what type of log is required for that use case.
Bad Query/System Performance
In case of bad performance of a query or complete system, Support needs:
- SQL/Server logs for before(if available) and after the performance issue.
- Potential profiling of the problematic query.
- The comparison session IDs.
- The ClusterOS logs.
In case of transaction conflicts, Support needs the SQL/Server logs with named Session ID to resolve the problem.
SQL Process Aborted
In case of aborted SQL processes, Support needs:
- SQL/Server Processes
In case of import or JDBC issues, Support needs:
- SQL/Server for the entire day
- JDBC backtraces
In case of database crash, Support needs:
- SQL/Server Processes
- ClusterOS logs
The following sections describe how to get specific logs with debug information: