[
https://issues.apache.org/jira/browse/CASSANDRA-10585?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15008933#comment-15008933
]
Carl Yeksigian commented on CASSANDRA-10585:
--------------------------------------------
While the version for 2.1 is fine, for 2.2 and 3.0, I don't think we should
switch to {{ExponentiallyDecayingReservoir}}; the reasoning behind us using
{{EstimatedHistogram}} is discussed in CASSANDRA-5657.
I'd like to either come up with a way to either estimate the statistic trying
to be calculated here, or extend {{EstimatedHistogram}} to properly capture a 0
value. I think we can approximate this by using the
{{RowCacheHit}}/{{RowCacheMiss}} metrics; that would mean
{{SSTablesPerReadHistogram}} is only captured on {{RowCacheMiss}}, so the new
metric might be
{noformat}
(RowCacheMiss/(RowCacheHit + RowCacheMiss)) * SSTablesPerRead
{noformat}
Do either of those seem feasible, [~isburmistrov]?
> SSTablesPerReadHistogram seems wrong when row cache hit happend
> ---------------------------------------------------------------
>
> Key: CASSANDRA-10585
> URL: https://issues.apache.org/jira/browse/CASSANDRA-10585
> Project: Cassandra
> Issue Type: Bug
> Reporter: Ivan Burmistrov
> Priority: Minor
> Fix For: 2.1.x, 2.2.x, 3.0.x
>
> Attachments: SSTablePerReadHistogram_RowCache-cassandra-2_1.patch,
> SSTablePerReadHistogram_RowCache-cassandra-2_2.patch,
> SSTablePerReadHistogram_RowCache-cassandra-3_0.patch
>
>
> SSTablePerReadHistogram metric now not considers case when row has been read
> from row cache.
> And so, this metric will have big values even almost all requests processed
> by row cache (and without touching SSTables, of course).
> So, it seems that correct behavior is to consider that if we read row from
> row cache then we read zero SSTables by this request.
> The patch at the attachment.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
