[
https://issues.apache.org/jira/browse/CASSANDRA-19391?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17817452#comment-17817452
]
Marcus Eriksson edited comment on CASSANDRA-19391 at 2/14/24 5:06 PM:
----------------------------------------------------------------------
flushing showed that we couldn't really read the metadata_snapshots sstables
due to the reversed longtoken localpartitioner we added in CASSANDRA-19189, so
here we add a reverse ordered partitioner (for long keys) which calculates
tokens by Long.MAX_VALUE - key.
CI a bit shaky, but looks like unrelated failures, will rerun (includes both
CASSANDRA-19390 and CASSANDRA-19391)
was (Author: krummas):
flushing showed that we couldn't really read the metadata_snapshots sstables
due to the reversed longtoken partitioner we added in CASSANDRA-19189, so here
we add a reverse ordered partitioner (for long keys) which calculates tokens by
Long.MAX_VALUE - key.
CI a bit shaky, but looks like unrelated failures, will rerun (includes both
CASSANDRA-19390 and CASSANDRA-19391)
> Flush metadata snapshot table on every write
> --------------------------------------------
>
> Key: CASSANDRA-19391
> URL: https://issues.apache.org/jira/browse/CASSANDRA-19391
> Project: Cassandra
> Issue Type: Improvement
> Components: Transactional Cluster Metadata
> Reporter: Marcus Eriksson
> Assignee: Marcus Eriksson
> Priority: Low
> Fix For: 5.x
>
> Attachments: ci_summary.html, result_details.tar.gz
>
>
> We depend on the latest snapshot when starting up, flushing avoids gaps
> between latest snapshot and the most recent local log entry
--
This message was sent by Atlassian Jira
(v8.20.10#820010)
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
