[
https://issues.apache.org/jira/browse/PHOENIX-5494?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16978114#comment-16978114
]
chenglei edited comment on PHOENIX-5494 at 11/20/19 6:54 AM:
-------------------------------------------------------------
[~kozdemir] [~larsh], thank you for the review.
[~kozdemir], I looked into the failed test
{{ConcurrentMutationsIT.testDeleteRowAndUpsertValueAtSameTS1}} after apply
{code:java}
else {
scan.setRaw(false);
scan.setMaxVersions(1);
}
{code}
It seems that because we set the {{scan.setRaw(false)}}, the delete marker
cell could not be scanned, so the put takes effect. I think although we can not
set {{scan.setRaw(false)}}, but we can set to retrieve the latest timestamp
cell for regular write, so I make a modification according to your suggestion:
{code:java}
SkipScanFilter skipScanFilter = scanRanges.getSkipScanFilter();
if(indexMetaData.getReplayWrite() != null) {
long timestamp =
getMaxTimestamp(dataTableMutationsWithSameRowKeyAndTimestamp);
scan.setTimeRange(0, timestamp);
scan.setFilter(new SkipScanFilter(skipScanFilter, true));
} else {
scan.setMaxVersions(1);
scan.setFilter(skipScanFilter);
}
{code}
The tests are passed.
I Uploaded my newest PHOENIX-5494_v5-4.x-HBase-1.4.patch
was (Author: comnetwork):
[~kozdemir] [~larsh], thank you for the review.
[~kozdemir], I looked into the failed test
{{ConcurrentMutationsIT.testDeleteRowAndUpsertValueAtSameTS1}} after apply
{code:java}
else {
scan.setRaw(false);
scan.setMaxVersions(1);
}
{code}
It seems that because we set the {{scan.setRaw(false)}}, the delete marker
cell could not be scanned, so the put takes effect. I think although we can not
set {{scan.setRaw(false)}}, but we can set to retrieve the latest timestamp
cell for regular write, so I make a modification according to your suggestion:
{code:java}
SkipScanFilter skipScanFilter = scanRanges.getSkipScanFilter();
if(indexMetaData.getReplayWrite() != null) {
long timestamp =
getMaxTimestamp(dataTableMutationsWithSameRowKeyAndTimestamp);
scan.setTimeRange(0, timestamp);
scan.setFilter(new SkipScanFilter(skipScanFilter, true));
} else {
scan.setMaxVersions(1);
scan.setFilter(skipScanFilter);
}
{code}
The tests are passed.
> Batched, mutable Index updates are unnecessarily run one-by-one
> ---------------------------------------------------------------
>
> Key: PHOENIX-5494
> URL: https://issues.apache.org/jira/browse/PHOENIX-5494
> Project: Phoenix
> Issue Type: Improvement
> Reporter: Lars Hofhansl
> Assignee: Kadir OZDEMIR
> Priority: Major
> Labels: performance
> Attachments: 5494-4.x-HBase-1.5.txt,
> PHOENIX-5494-4.x-HBase-1.4.patch, PHOENIX-5494.master.001.patch,
> PHOENIX-5494.master.002.patch, PHOENIX-5494.master.003.patch,
> PHOENIX-5494_v3-4.x-HBase-1.4.patch, PHOENIX-5494_v5-4.x-HBase-1.4.patch,
> Screenshot_20191110_160243.png, Screenshot_20191110_160351.png,
> Screenshot_20191110_161453.png
>
> Time Spent: 2h 10m
> Remaining Estimate: 0h
>
> I just noticed that index updates on mutable tables retrieve their deletes
> (to invalidate the old index entry) one-by-one.
> For batches, this can be *the* major time spent during an index update. The
> cost is mostly incured by the repeated setup (and seeking) of the new region
> scanner (for each row).
> We can instead do a skip scan and get all updates in a single scan per region.
> (Logically that is simple, but it will require some refactoring)
> I won't be getting to this, but recording it here in case someone feels
> inclined.
--
This message was sent by Atlassian Jira
(v8.3.4#803005)
