Misleading error message after database restart/failover
Details
| Detail name | Value |
|---|---|
| Changelog Number | 10258 |
| Type | Bug |
| Status | Resolved |
| Affected Versions | EXASOL 6.0.0, Exasol 6.1.0, Exasol 6.2.0, Exasol 7.0.alpha1 |
| Fix Versions | Exasol 6.1.10, Exasol 6.2.8, Exasol 7.0.rc1 |
| Resolution Date | 2020-06-24 |
In order to ensure consistent performance and data redundancy after a cluster node has failed and was replaced by a reserve node, the database data that was stored on the failed node is usually moved to the new active node automatically (former reserve node).
In cases where there were multiple node failures before this process could be completed, manual intervention is currently required to ensure that each database process is running on the correct node.
In some scenarios where multiple node failures occurred, a misleading internal error message was shown:
Could not automatically move data of database system <db_name>: Failed to move nodes: permission denied
instead of the intended message.
Automatically move offline node n0011 for volume 0 of database <db_name> to online node n0012 failed, because it is already in use, manual intervention necessary!'
Two steps should be performed in this case
- Check the storage volume and move data from any nodes that are expected to be unavailable for the foreseeable future. You can find instructions to do this here: https://docs.exasol.com/administration/on-premise/nodes/replace_active_node.htm#MoveDataSegmentstoTargetNode
- Restart the database so each database process is able to access its data locally.
The error message will be changed to the intended message:
Automatically move offline node n0011 for volume 0 of database <db_name> to online node n0012 failed, because it is already in use, manual intervention necessary!'
