Hi all,
We are using elasticsearch 1.3.7 cluster with five data only nodes (heap 6GB) and three master only nodes (heap 4GB).
FYI, Upgrading ES version is on progress
We got node removal from a cluster for a 1.9 minutes , so while check with log and found that long pause GC is occurs and three slow queries are reported.
While check with Slow query, three queries are same and with size of 10000000.
{"size":10000000,"query":{"bool":{"must":{"term":{"obj_id":545653}}}}}
{"size":10000000,"query":{"bool":{"must":{"term":{"obj_id":545653}}}}}
{"size":10000000,"query":{"bool":{"must":{"term":{"obj_id":545653}}}}}
But same time, before these three queries came Heap used level is fine ( 3.8 GB is used out of 6 GB).after these queries came Long pause GC is occurs.
Checked in ES documents, "obj_id":545653 in that specified type there is only one document is alone found.
1.How come getting one record as result is trigerring GC?
2.whether size ("size":10000000,) is reason for GC?
3. is there is any way to avoid long pause GC in ES?
Can you please debug and help on this