[
https://issues.apache.org/jira/browse/CASSANDRA-2698?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13689804#comment-13689804
]
Benedict commented on CASSANDRA-2698:
-------------------------------------
Hi Yuki,
{quote}
how many rows each merkle tree range account for (and the size that this
represent)
{quote}
I take it two histograms, one of row distribution, the other of size (of row)
distribution, is what you're looking for?
I've attached a new patch which is *not* complete, in that I have not tested it
and may want to change a few of the final details (such as, possibly, where and
what is logged with the histogram), but before I iron out those kinks I wanted
to run past the main crux of the changes to see if it's what you're looking
for. Simply put, the merkle tree ranges now retain both a sizeOfRange (ie size
of rows added) and rowsInRange (ie number of rows added). Merkle tree now
exposes two histogramXXX() methods which use these, and which as of now are
logged in Validator.complete(). As it stands, I serialise both these new values
over the wire with any merkle tree, to ensure no unexpected behaviour for
future users of the class, and as such I also retained my TreeDifference
changes to the merkle tree, which reports the size and row count of each side
of a difference. These latter two changes may be slightly controversial, so
want to run them past you, as well as confirm the basic information I'm
printing is what you're looking for.
Cheers!
> Instrument repair to be able to assess it's efficiency (precision)
> ------------------------------------------------------------------
>
> Key: CASSANDRA-2698
> URL: https://issues.apache.org/jira/browse/CASSANDRA-2698
> Project: Cassandra
> Issue Type: Improvement
> Reporter: Sylvain Lebresne
> Assignee: Benedict
> Priority: Minor
> Labels: lhf
> Attachments: nodetool_repair_and_cfhistogram.tar.gz,
> patch_2698_v1.txt, patch.diff, patch-rebased.diff
>
>
> Some reports indicate that repair sometime transfer huge amounts of data. One
> hypothesis is that the merkle tree precision may deteriorate too much at some
> data size. To check this hypothesis, it would be reasonably to gather
> statistic during the merkle tree building of how many rows each merkle tree
> range account for (and the size that this represent). It is probably an
> interesting statistic to have anyway.
--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira
