During extremely high load the cluster may become unstable
Details
| Detail name | Value |
|---|---|
| Changelog Number | 27901 |
| Type | Improvement |
| Status | Resolved |
| Fix Versions | Exasol 2025.2.0, Exasol 2025.1.9 |
| Resolution Date | 2025-12-17 |
Description
Under heavy load, where a lot of messages and events over the network may be sent, the cluster will continuously reconfigure itself. As a result, the database may become unresponsive and crash. Increasing volume redundancy under high load has been identified as a trigger for this behavior.
Improvement
The protocols were improved so that the cluster will remain stable, even under substantial load and triggering of a volume redundancy increase. However, some interfaces (like csinfo or cosps) may experience timeouts.
