Replies: 2 comments 5 replies
-
Hi! Probably you may take a look at cpu/mem/disk etc. usage during the slow query period to see if there is any evidence? Basically check the overall ClickHouse cluster metrics to see any factor that occupied some resources. |
Beta Was this translation helpful? Give feedback.
1 reply
-
Clickhouse version? |
Beta Was this translation helpful? Give feedback.
4 replies
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
-
As shown below, the same SQL query was executed twice, but it was actually executed many times. I only display the logs for these two executions. The one at 10:11:31 took over 75 seconds, while the one which was earier at 10:10:30 only took 884milliseconds
I compared the ProfileEvents field. The following image may provide a more intuitive comparison
I found that apart from the NetworkReceiveElapsedMicroseconds metric, the other metrics don't differ much.
However, since the unit of this metric is microseconds, it doesn't seem to account for such a large difference. I'm stuck and haven't found the cause of the problem. Can someone help me take a look, or provide me with a correct troubleshooting approach? I can provide additional information if needed. Thank you very much.
Additionally, I'd like to mention that this slow SQL issue occurs sporadically.
Beta Was this translation helpful? Give feedback.
All reactions