[
https://issues.apache.org/jira/browse/HBASE-20312?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16443632#comment-16443632
]
Yu Li commented on HBASE-20312:
-------------------------------
[~chancelq] I think Mike's suggestion of creating several sub-tasks to make
review easier makes sense. For example, we could have below tasks (or maybe
more, let's discuss):
1. Introducing CCSMap
- The basic, compacted data structure that is more GC friendly than jdk's
CSLM implementation
2. Introducing CCSMapMemstore
- Integration with HBase
[~mdrob] Yes we did think about making the CCSMap into guava-like project, but
on one hand the CCSMap requires key/value to be serializable so we're not sure
whether guava-like project would like to accept it, and on the other hand this
(adding dependency to upstreaming to other project) will slow down the
contribution to our hbase. Maybe we could follow the same pattern as our
asynchronous hdfs output stream implementation
{{FanOutOneBlockAsyncDFSOutput}}, that we introduce it into hbase first, and
later on if we could make it into HDFS project, we could change to refer to
that. What do you think? Thanks.
> CCSMap: A faster, GC-friendly, less memory Concurrent Map for memstore
> ----------------------------------------------------------------------
>
> Key: HBASE-20312
> URL: https://issues.apache.org/jira/browse/HBASE-20312
> Project: HBase
> Issue Type: New Feature
> Components: regionserver
> Reporter: Xiang Wang
> Assignee: Chance Li
> Priority: Major
> Fix For: 3.0.0
>
> Attachments: 1.1.2-ccsmap-number.png, HBASE-20312-1.3.2.patch,
> HBASE-20312-master.v1.patch, HBASE-20312-master.v2.patch,
> HBASE-20312-master.v3.patch, HBASE-20312-master.v4.patch,
> HBASE-20312-master.v5.patch, HBASE-20312-master.v6.patch,
> HBASE-20312-master.v7.patch, ccsmap-branch-1.1.patch, hits.png, jira1.png,
> jira2.png, jira3.png, off-heap-test-put-master.png,
> on-heap-test-put-master.png
>
>
> Now hbase use ConcurrentSkipListMap as memstore's data structure.
> Although MemSLAB reduces memory fragment brought by key-value pairs.
> Hundred of millions key-value pairs still make young generation
> garbage-collection(gc) stop time long.
>
> These are 2 gc problems of ConcurrentSkipListMap:
> 1. HBase needs 3 objects to store one key-value on expectation. One
> Index(skiplist's average node height is 1), one Node, and one KeyValue. Too
> many objects are created for memstore.
> 2. Recent inserted KeyValue and its map structure(Index, Node) are assigned
> on young generation.The card table (for CMS gc algorithm) or RSet(for G1 gc
> algorithm) will change frequently on high writing throughput, which makes YGC
> slow.
>
> We devleoped a new skip-list map called CompactdConcurrentSkipListMap(CCSMap
> for short),
> which provides similary features like ConcurrentSkipListMap but get rid of
> Objects for every key-value pairs.
> CCSMap's memory structure is like this picture:
> !jira1.png!
>
> One CCSMap consists of a certain number of Chunks. One Chunk consists of a
> certain number of nodes. One node is corresspding one element. This element's
> all information and its key-value is encoded on a continuous memory segment
> without any objects.
> Features:
> 1. all insert,update,delete operations is lock-free on CCSMap.
> 2. Consume less memory, it brings 40% memory saving for 50Byte key-value.
> 3. Faster on small key-value because of better cacheline usage. 20~30% better
> read/write troughput than ConcurrentSkipListMap for 50Byte key-value.
> CCSMap do not support recyle space when deleting element. But it doesn't
> matter for hbase because of region flush.
> CCSMap has been running on Alibaba's hbase clusters over 17 months, it cuts
> down YGC time significantly. here are 2 graph of before and after.
> !jira2.png!
> !jira3.png!
>
>
--
This message was sent by Atlassian JIRA
(v7.6.3#76005)
