[
https://issues.apache.org/jira/browse/CASSANDRA-9060?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14386320#comment-14386320
]
Gustav Munkby commented on CASSANDRA-9060:
------------------------------------------
I might be horribly mistaken, but my reading of estimatedKeysForRanges in
SSTableReader is that it is based on the number of keys that are definitely in
the range, given the sample in the index summary. Thus the estimate should be
guaranteed to be an under-approximation.
If my understanding above is correct, it seems the repaired table would
typically have a too small Bloom filter with the added tweak. Similarly, the
unrepaired table will typically have a slightly too big Bloom filter. Given
that the Bloom filter is only an optimisation, I'm not sure either of those
things really matter that much. I guess it depends on whether any other pieces
of the code assume the Bloom-filter sizes to be over- or under sized.
> Anticompaction hangs on bloom filter bitset serialization
> ----------------------------------------------------------
>
> Key: CASSANDRA-9060
> URL: https://issues.apache.org/jira/browse/CASSANDRA-9060
> Project: Cassandra
> Issue Type: Bug
> Reporter: Gustav Munkby
> Assignee: Gustav Munkby
> Priority: Minor
> Fix For: 2.1.4
>
> Attachments: 0001-another-tweak-to-9060.patch, 2.1-9060-simple.patch,
> trunk-9060.patch
>
>
> I tried running an incremental repair against a 15-node vnode-cluster with
> roughly 500GB data running on 2.1.3-SNAPSHOT, without performing the
> suggested migration steps. I manually chose a small range for the repair
> (using --start/end-token). The actual repair part took almost no time at all,
> but the anticompactions took a lot of time (not surprisingly).
> Obviously, this might not be the ideal way to run incremental repairs, but I
> wanted to look into what made the whole process so slow. The results were
> rather surprising. The majority of the time was spent serializing bloom
> filters.
> The reason seemed to be two-fold. First, the bloom-filters generated were
> huge (probably because the original SSTables were large). With a proper
> migration to incremental repairs, I'm guessing this would not happen.
> Secondly, however, the bloom filters were being written to the output one
> byte at a time (with quite a few type-conversions on the way) to transform
> the little-endian in-memory representation to the big-endian on-disk
> representation.
> I have implemented a solution where big-endian is used in-memory as well as
> on-disk, which obviously makes de-/serialization much, much faster. This
> introduces some slight overhead when checking the bloom filter, but I can't
> see how that would be problematic. An obvious alternative would be to still
> perform the serialization/deserialization using a byte array, but perform the
> byte-order swap there.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
