-
Notifications
You must be signed in to change notification settings - Fork 1.4k
nodetool cfstats hangs forever #5414
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Comments
@amoskong so its clear:
|
It was after bench was done to generate the load. When I run nodetool cfstats, zero load in the cluster. |
@amnonh can you please start the debugging and indicate where its stuck on scylla end |
It didn't reproduce for me, it does take a long time, on my laptop it took 25s |
cosing the issue if it reproduces - please open a new bug with exact time it took and how much time you waited (and if there were stalls in scylla) |
Same issue here, after I did
If operation is legitimately should be taking a long time it would be nice to at least get some hint/warning from nodetool... |
@let4be - agreed, yet to understand if its long time or hangged can you please share the info on how long you waited for nodetool cfstats to return. |
I waited over 20 min with zero success - process When all this failed I completely took pressure off scylla and stopped all data producers, scylla cpu usage fell basically to zero but I still could not get a response from When it works I'm able to get (running single node scylla on very good hardware) |
@amnonh ping |
I'm still unable to reproduce. @let4be is is something you can reproduce easily? If so, can you run scylla-jmx with logging enabled:
|
no response from users - if this happens - please provide requested information #5414 (comment) |
reproduced in 4.4 2TB longevity during showTopPartitions nemesis. nodetool cfstats got stuck, executing on show-top-partitions nemesis. the nodetool cfstats commands executed until got stuck are:
the full nemesis list before nodetool got stuck is:
Installation details Test: Issue description ==================================== PUT ISSUE DESCRIPTION HERE ==================================== Restore Monitor Stack command: Test id: Logs: |
according to @roydahan this issue can be re-closed since there's a newer open issue for that, (not sure which one). |
Scylla: 3.1.2
Nodetool: 35906df
Start two 2 nodes in the cluster
Inject data with:
$ scylla-bench -workload sequential -mode write -nodes 127.0.0.1 -rows-per-request 1 -partition-count 1000 -clustering-row-count 100 -clustering-row-size 1024000 -keyspace myks1024000 -table t1 -replication-factor 2
Then run nodetool. It never returns
$ nodetool cfstats
Other nodetool cmds work correctly:
The text was updated successfully, but these errors were encountered: