Local plan for parallel replicas#64448
Conversation
|
This is an automated comment for commit c53e165 with description of existing statuses. It's updated for the latest CI running ✅ Click here to open a full report in a separate page Successful checks
|
|
Waiting for #69264 |
| executeQueryWithParallelReplicasCustomKey(query_plan, storage_id, query_info, columns, snapshot, processed_stage, header, context); | ||
| } | ||
|
|
||
| bool canUseParallelReplicasOnInitiator(const ContextPtr & context) |
There was a problem hiding this comment.
Could I kindly inquire about a matter? Why make a new function here to check if a cluster (shard) has only one node, rather than within the Context::canUseParallelReplicasOnInitiator()? If a cluster (shard) has only one node, all ParallelReplicas scenarios need to fall back to local execution, not just in buildQueryPlanForTableExpression(). @devcrafter Looking forward to your reply.
There was a problem hiding this comment.
Hi @zoomxi, it makes sense. If you're up to it, please feel free to provide a PR. Probably, a couple of requirements for such a case, - there should be (1) a warning in traces about fallback to local execution, and (2) a test to check this behavior.
Local plan for parallel replicas
Local plan for parallel replicas

Changelog category (leave one):
CI Settings
NOTE: If your merge the PR with modified CI you MUST KNOW what you are doing
NOTE: Checked options will be applied if set before CI RunConfig/PrepareRunConfig step