scylla-cluster-tests
scylla-cluster-tests copied to clipboard
Alter table with TWCS with large default TTL return misconfiguration cqlsh error
Prerequisites
- [x] Are you rebased to master ?
- [x] Is it reproducible ?
- [x] Did you perform a cursory search if this issue isn't opened ?
Versions
- SCT: master
- scylla: master
Logs
- test_id: a84c3f29-6685-4274-a95f-139360b35924
- job log: https://cloudius-jenkins-test.s3.amazonaws.com/a84c3f29-6685-4274-a95f-139360b35924/20230101_054449/sct-runner-a84c3f29.tar.gz
Description
if table has TWCS or was changed with modify_table nemesis to TWCS, next nemesis modify table which try to alter default_time_to_live and compaction_window could fail with error:
cassandra.protocol.ConfigurationException: <Error from server: code=2300 [Query invalid because of configuration issue] message="The setting of default_time_to_live=16842639 and compaction window=86400(s) can lead to 194 windows, which is larger than the allowed number of windows specified by the twcs_max_window_count (50) parameter. Note that default_time_to_live=0 is also highly discouraged.">
it is not an error , it is protection from misconfiguration for TWCS to avoid large number of sstables.
need to add to sct fix, which will not stop nemesis with error:
- check error message and mark nemesis as skipped, or
- add check to _modify_default_time_to_live if table has TWCS and get compaction_window and correctly calculate numbers
Steps to Reproduce
- Create table or alter table with TWCS 2.modify_table with default_time_to_live large number
- get error on cqlsh command execution
Expected behavior: [What you expected to happen]
cqlsh error: cassandra.protocol.ConfigurationException: <Error from server: code=2300 [Query invalid because of configuration issue] message="The setting of default_time_to_live=16842639 and compaction window=86400(s) can lead to 194 windows, which is larger than the allowed number of windows specified by the twcs_max_window_count (50) parameter. Note that default_time_to_live=0 is also highly discouraged.">
@aleksbykov option 2, to calculate the correct configuration,
we shouldn't be skipping after the fact, and parsing error messages..
any progress with this one?
Put it to queue
This issue is stale because it has been open 2 years with no activity. Remove stale label or comment or this will be closed in 2 days.
This issue was closed because it has been stalled for 2 days with no activity.