This is a fix release that includes the following changes:
- Fixed handling of NULL values in the cluster tab of the Admin UI.
- Updated the Size label in the table detail tab of the Admin UI to Size (Sum of primary shards) for clarification.
- Fixed an issue where ordering by a
TIMESTAMPscalar would cause the query to stick.
- Fixed a potential
SHOW CREATE TABLEwith plain type indices.
- Fixed an issue causing queries ordered by a
STRINGscalar with possible
NULLvalues to stick.
- Fixed an issue causing nodes to crash with
OutOfMemoryExceptionwhen executing queries with
ORDER BYon large data sets were executed. Now the query circuit breaker will be tripped causing the query to be killed with the node’s health unaffected.
- Improved the error handling for a race condition to ensure it doesn’t cause queries to fail to return a value, essentially moving to a stuck state.
- Improved the behavior handling when a cluster drops below the number of minimum nodes. The root REST endpoint will no longer timeout and
sys.checkscan still be queried, similar to how
sys.nodescan still be queried.As with all releases, be sure to back up your data prior to upgrading.
We look forward to bringing more features and more fixes to CrateDB! Stay tuned for further details!