cockroachdb / cockroachdb/cockroach
In the sysbench oltp_point_select scenario, when running with 400 concurrent connections, there is a sharp drop in QPS (Queries Per Second), and the CPU remains underutilized.
- Dominant language
- Go
- Stars
- 32.5k
- Forks
- 4.1k
- PR merge metrics
- PR metrics pending
Description
**Describe the problem**
In the sysbench oltp_point_select scenario, when running with 400 concurrent connections, there is a sharp drop in QPS (Queries Per Second), and the CPU remains underutilized.
**To Reproduce**
What did you do? Describe in your own words.
test on crdb cloud
instance 16c64g *3 with 500g storage
verion 23.1.8
command:
sysbench oltp_point_select --db-driver=pgsql --report-interval=10 --table-size=10000000 --tables=32 --pgsql-host=xxx--pgsql-port=26257 --pgsql-user=xxx--pgsql-db=test --pgsql-sslmode=verify-ca --pgsql-password='xxxxxx' --time=1200 --auto-inc=false --threads=400 run
Jira issue: CRDB-31285
Contributor guide
Assessment
This issue has not been assessed yet.