[
https://issues.apache.org/jira/browse/CASSANDRA-21085?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Andy Salnikov updated CASSANDRA-21085:
--------------------------------------
Description:
I have upgraded 4.1.9 cluster to 5.0.6 and I see confusing warnings in the
server log which say {{{}Failed trySkipCache on file: ... Error: Invalid
argument{}}}. The file name in the warning seems to refer to a file that is
being compacted. Here is an example of messages from the log that refer to one
such file:
{noformat}
INFO [CompactionExecutor:14] 2025-12-11 00:12:58,479 CompactionTask.java:175 -
Compacting (261977f0-d626-11f0-b282-6b53d56b692b) [
/var/lib/cassandra/<table-dir>/nb-12567-big-Data.db:level=0,
/var/lib/cassandra/<table-dir>/nb-13955-big-Data.db:level=0,
/var/lib/cassandra/<table-dir>/nb-16807-big-Data.db:level=0,
/var/lib/cassandra/<table-dir>/nb-15350-big-Data.db:level=0, ]
WARN [NonPeriodicTasks:1] 2025-12-11 00:22:55,106 NoSpamLogger.java:107 -
Failed trySkipCache on file:
/var/lib/cassandra/<table-dir>/nb-13955-big-Data.db Error: Invalid argument
INFO [NonPeriodicTasks:1] 2025-12-11 00:30:05,808 BigFormat.java:231 -
Deleting sstable: /opt/cassandra/data/<table-dir>/nb-13955-big
{noformat}
I interpret this so that compaction for nb-13955-big-Data.db ran from 00:12:58
to 00:30:05, but trySkipCache accessed that file during compaction and failed?
And I saw the same warnings when running {{{}stableupgrade{}}}, it does not
happen for every table, likely depends on holw long does it take to upgrade:
{noformat}
$ sstableupgrade keyspace table
Found 4 sstables that need upgrading.
Upgrading
BigTableReader:big(path='/var/lib/cassandra/data/keyspace/table-4053f9506dc711f08959f1716794a923/nb-3121-big-Data.db')
WARN 19:07:36,477 Failed trySkipCache on file:
/var/lib/cassandra/data/keyspace/table-4053f9506dc711f08959f1716794a923/nb-3121-big-Data.db
Error: Invalid argument
Upgrade of
BigTableReader:big(path='/var/lib/cassandra/data/keyspace/table-4053f9506dc711f08959f1716794a923/nb-3121-big-Data.db')
complete.
{noformat}
And I saw this warning to happen more than once per table for long upgrades,
warnings come 10 minutes apart.
Cassandra and sstableupgrade run in a container using a standard image from
DockerHub, version 5.0.6. Host OS is Rocky Linux 9.6.
Thread on Slack:
[https://the-asf.slack.com/archives/CJZLTM05A/p1765415636431169]
was:
I have upgraded 4.1.9 cluster to 5.0.6 and I see confusing warnings in the
server log which say {{Failed trySkipCache on file: ... Error: Invalid
argument}}. The file name in the warning seems to refer to a file that is
being compacted. Here is an example of messages from the log that refer to one
such file:
{noformat}
INFO [CompactionExecutor:14] 2025-12-11 00:12:58,479 CompactionTask.java:175 -
Compacting (261977f0-d626-11f0-b282-6b53d56b692b) [
/var/lib/cassandra/<table-dir>/nb-12567-big-Data.db:level=0,
/var/lib/cassandra/<table-dir>/nb-13955-big-Data.db:level=0,
/var/lib/cassandra/<table-dir>/nb-16807-big-Data.db:level=0,
/var/lib/cassandra/<table-dir>/nb-15350-big-Data.db:level=0, ]
WARN [NonPeriodicTasks:1] 2025-12-11 00:22:55,106 NoSpamLogger.java:107 -
Failed trySkipCache on file:
/var/lib/cassandra/<table-dir>/nb-13955-big-Data.db Error: Invalid argument
INFO [NonPeriodicTasks:1] 2025-12-11 00:30:05,808 BigFormat.java:231 -
Deleting sstable: /opt/cassandra/data/<table-dir>/nb-13955-big
{noformat}
I interpret this so that compaction for nb-13955-big-Data.db ran from 00:12:58
to 00:30:05, but trySkipCache accessed that file during compaction and failed?
And I saw the same warnings when running {{stableupgrade}}, it does not happen
for every table, likely depends on holw long does it take to upgrade:
{noformat}
$ sstableupgrade keyspace table
Found 4 sstables that need upgrading.
Upgrading
BigTableReader:big(path='/var/lib/cassandra/data/keyspace/table-4053f9506dc711f08959f1716794a923/nb-3121-big-Data.db')
WARN 19:07:36,477 Failed trySkipCache on file:
/var/lib/cassandra/data/keyspace/table-4053f9506dc711f08959f1716794a923/nb-3121-big-Data.db
Error: Invalid argument
Upgrade of
BigTableReader:big(path='/var/lib/cassandra/data/keyspace/table-4053f9506dc711f08959f1716794a923/nb-3121-big-Data.db')
complete.
{noformat}
Cassandra and sstableupgrade run in a container using a standard image from
DockerHub, version 5.0.6. Host OS is Rocky Linux 9.6.
Thread on Slack: https://the-asf.slack.com/archives/CJZLTM05A/p1765415636431169
> Warning about failed trySkipCache in Cassandra 5.0
> --------------------------------------------------
>
> Key: CASSANDRA-21085
> URL: https://issues.apache.org/jira/browse/CASSANDRA-21085
> Project: Apache Cassandra
> Issue Type: Bug
> Components: Local/Compaction
> Reporter: Andy Salnikov
> Priority: Normal
>
> I have upgraded 4.1.9 cluster to 5.0.6 and I see confusing warnings in the
> server log which say {{{}Failed trySkipCache on file: ... Error: Invalid
> argument{}}}. The file name in the warning seems to refer to a file that is
> being compacted. Here is an example of messages from the log that refer to
> one such file:
> {noformat}
> INFO [CompactionExecutor:14] 2025-12-11 00:12:58,479 CompactionTask.java:175
> -
> Compacting (261977f0-d626-11f0-b282-6b53d56b692b) [
> /var/lib/cassandra/<table-dir>/nb-12567-big-Data.db:level=0,
> /var/lib/cassandra/<table-dir>/nb-13955-big-Data.db:level=0,
> /var/lib/cassandra/<table-dir>/nb-16807-big-Data.db:level=0,
> /var/lib/cassandra/<table-dir>/nb-15350-big-Data.db:level=0, ]
> WARN [NonPeriodicTasks:1] 2025-12-11 00:22:55,106 NoSpamLogger.java:107 -
> Failed trySkipCache on file:
> /var/lib/cassandra/<table-dir>/nb-13955-big-Data.db Error: Invalid argument
> INFO [NonPeriodicTasks:1] 2025-12-11 00:30:05,808 BigFormat.java:231 -
> Deleting sstable: /opt/cassandra/data/<table-dir>/nb-13955-big
> {noformat}
> I interpret this so that compaction for nb-13955-big-Data.db ran from
> 00:12:58 to 00:30:05, but trySkipCache accessed that file during compaction
> and failed?
> And I saw the same warnings when running {{{}stableupgrade{}}}, it does not
> happen for every table, likely depends on holw long does it take to upgrade:
> {noformat}
> $ sstableupgrade keyspace table
> Found 4 sstables that need upgrading.
> Upgrading
> BigTableReader:big(path='/var/lib/cassandra/data/keyspace/table-4053f9506dc711f08959f1716794a923/nb-3121-big-Data.db')
> WARN 19:07:36,477 Failed trySkipCache on file:
> /var/lib/cassandra/data/keyspace/table-4053f9506dc711f08959f1716794a923/nb-3121-big-Data.db
> Error: Invalid argument
> Upgrade of
> BigTableReader:big(path='/var/lib/cassandra/data/keyspace/table-4053f9506dc711f08959f1716794a923/nb-3121-big-Data.db')
> complete.
> {noformat}
> And I saw this warning to happen more than once per table for long upgrades,
> warnings come 10 minutes apart.
> Cassandra and sstableupgrade run in a container using a standard image from
> DockerHub, version 5.0.6. Host OS is Rocky Linux 9.6.
> Thread on Slack:
> [https://the-asf.slack.com/archives/CJZLTM05A/p1765415636431169]
--
This message was sent by Atlassian Jira
(v8.20.10#820010)
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
