hbase-issues mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "ramkrishna.s.vasudevan (JIRA)" <j...@apache.org>
Subject [jira] [Updated] (HBASE-15871) Memstore flush doesn't finish because of backwardseek() in memstore scanner.
Date Wed, 24 Aug 2016 10:58:20 GMT

     [ https://issues.apache.org/jira/browse/HBASE-15871?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel

ramkrishna.s.vasudevan updated HBASE-15871:
    Attachment: HBASE-15871.patch

Patch for trunk. This unnecessary backward seek can also be seen in trunk. But in lower branches
this could be more significant because we hold a lock too. In trunk the lock is removed.
How ever I think [~Jeongdae Kim]'s first patch is good as I am not sure if a snapshot fails
we do retry and even if that fails we throw DroppedSnapshotException. Do we try to reuse that
snapshot? Also during the retry process the seqId to be generated by this flush is fixed and
I think for the current store there cannot be concurrent flushes till the snapshot is cleared.

This patch also adds a test case. Ping for reviews/feedback. It seems to be a critical fix
for any user using 'reverse' scan. should be a perf fix too. 

> Memstore flush doesn't finish because of backwardseek() in memstore scanner.
> ----------------------------------------------------------------------------
>                 Key: HBASE-15871
>                 URL: https://issues.apache.org/jira/browse/HBASE-15871
>             Project: HBase
>          Issue Type: Bug
>          Components: Scanners
>    Affects Versions: 1.1.2
>            Reporter: Jeongdae Kim
>             Fix For: 1.1.2
>         Attachments: HBASE-15871.branch-1.1.001.patch, HBASE-15871.branch-1.1.002.patch,
HBASE-15871.branch-1.1.003.patch, HBASE-15871.patch, memstore_backwardSeek().PNG
> Sometimes in our production hbase cluster, it takes a long time to finish memstore flush.(
for about more than 30 minutes)
> the reason is that a memstore flusher thread calls StoreScanner.updateReaders(), waits
for acquiring a lock that store scanner holds in StoreScanner.next() and backwardseek() in
memstore scanner runs for a long time.
> I think that this condition could occur in reverse scan by the following process.
> 1) create a reversed store scanner by requesting a reverse scan.
> 2) flush a memstore in the same HStore.
> 3) puts a lot of cells in memstore and memstore is almost full.
> 4) call the reverse scanner.next() and re-create all scanners in this store because all
scanners was already closed by 2)'s flush() and backwardseek() with store's lastTop for all
new scanners.
> 5) in this status, memstore is almost full by 2) and all cells in memstore have sequenceID
greater than this scanner's readPoint because of 2)'s flush(). this condition causes searching
all cells in memstore, and seekToPreviousRow() repeatly seach cells that are already searched
if a row has one column. (described this in more detail in a attached file.)
> 6) flush a memstore again in the same HStore, and wait until 4-5) process finished, to
update store files in the same HStore after flusing.
> I searched HBase jira. and found a similar issue. (HBASE-14497) but, HBASE-14497's fix
can't solve this issue because that fix just changed recursive call to loop.(and already applied
to our HBase version)

This message was sent by Atlassian JIRA

View raw message