Bulk request is throwing rejected execution exception

{"reason":"rejected execution of processing of [6013644][indices:data/write/bulk[s][p]]: request: BulkShardRequest [[index-test24][3]] containing [52] requests, target allocation id: HwDviwdxRRexbw-sT7B8WQ, primary term: 3 on EsThreadPoolExecutor[name = one.cloud.com/write, queue capacity = 500, org.elasticsearch.common.util.concurrent.EsThreadPoolExecutor@1e05480d[Running, pool size = 32, active threads = 32, queued tasks = 500, completed tasks = 782262]]","type":"es_rejected_execution_exception"}

h/w: 6 nodes * 32 cores * 120 shards

heavily on parent-child relationship unto 4 levels deep with proper routing of the data by the parent ids, I'm seeing the above exception quiet often

but I've a different cluster of size 17 nodes * 80 shards which indexes just the straight time series data without any relationships (~ upto 40k/sec), I don't see any issues with the indexing and the cluster is also stable

Please suggest

If you have a reasonably large number of threads indexing into Elasticsearch in parallel and each bulk request contains a mix of routing IDs, this blog post might help explain the rejections.

This topic was automatically closed 28 days after the last reply. New replies are no longer allowed.