cassandra-lucene-index icon indicating copy to clipboard operation
cassandra-lucene-index copied to clipboard

Published 20 hours ago • verified publisher icon instaclustr

Performance concerns

Open Telavian opened this issue 5 years ago • 2 comments

I have a 3 node cluster I am using for testing. CPU: Intel(R) Xeon(R) CPU X5650 @ 2.67GHz Mem: 94GB Disk: SSD

Currently doing about 1000 updates a second seems to bring all the nodes to 100% CPU and start timing out operations. Is this normal?

What should I check or update?

Telavian avatar Sep 17 '20 21:09 Telavian

That sounds slow but I guess it depends on the size of the updates as well as could be some configuration issues. I would probably try doing the same inserts on a table without a lucene-based secondary index to see if it's lucene overhead or something else.

slater-ben avatar Sep 20 '20 05:09 slater-ben

These updates also were not just on the lucene indexed tables. Either the lucene indexes were bringing the whole cluster down or the non-lucene indexed tables were also slow for some reason.

Certainly switching back to elasticsearch caused the cluster to be very fast again. I know this information is very general and would love to pin it down to something more specific for you.

Do you have a load test currently?

Telavian avatar Sep 20 '20 06:09 Telavian