[
https://issues.apache.org/jira/browse/CASSANDRA-21832?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Alan Wang updated CASSANDRA-21832:
----------------------------------
Change Category: Semantic
Complexity: Low Hanging Fruit
Component/s: Local/SSTable
Status: Open (was: Triage Needed)
> Memory for an SSTable's index summary can be under allocated
> ------------------------------------------------------------
>
> Key: CASSANDRA-21832
> URL: https://issues.apache.org/jira/browse/CASSANDRA-21832
> Project: Apache Cassandra
> Issue Type: Improvement
> Components: Local/SSTable
> Reporter: Alan Wang
> Assignee: Alan Wang
> Priority: Normal
> Attachments: ci_summary.html, result_details.tar.gz
>
> Time Spent: 10m
> Remaining Estimate: 0h
>
> For SSTables using the BIG format, we use each SSTable's read rate to
> determine how much memory to allocate for a SSTable’s index summary. SSTables
> that are read more frequently get more memory. However, the metric used in
> determining this value undercounts as we do not factor in the times when a
> lookup is a false positive from the bloom filter.
--
This message was sent by Atlassian Jira
(v10.3.26#10030026)
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
