Network splits and suspended nodes may cause a node to remain offline after rejoining

Details

Detail name Value
Changelog Number 14913
Type Bug
Status Resolved
Affected Versions Exasol 8.0.0
Fix Versions Exasol 8.33.0
Resolution Date 2025-02-27

Description

When some nodes disconnect from the COS network and are marked as suspended (by an operator), they may remain as offline even after they are apparently online. This happens arbitrarily because it is a timing dependent issue.

Workaround

Force cluster re-synchronizations by running cosmod --force-synchronization.

Fix

The described scenario should not happen. Offline nodes that have been manually marked as suspended will appear online after rejoining the cluster.