[
https://issues.apache.org/jira/browse/ZOOKEEPER-3352?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16836096#comment-16836096
]
Fangmin Lv commented on ZOOKEEPER-3352:
---------------------------------------
Had an intern done some prototype with me by building the ZooKeeper on RocksDB,
the intention is to store more data, use incremental snapshot to reduce the
cost with LSM, reduce the GC effort and data loading time, etc. It also avoids
fuzzy snapshot which is pretty error-prone and caused lots of consistency bugs
in ZK. But it's likely won't give us the same throughput as what we have now
with in memory data tree and append only txn file.
Currently, we only moved snapshot to RocksDB, the plan is to remove the in
memory data tree as well, but only cache things like children structure,
metadata to retain txn prepare and read throughput, etc.
We tried to keep the file based txn file, but disable the WAL in RocksDB, since
we cannot see obvious benefit for now.
This is still a prototype, and we didn't have time in the last year to spend
more effort there, but it will be an interesting try to see how it goes when
ZooKeeper meet other backend.
[~belugabehr] what's the status of this project on your side, is it a thought
or you already have something?
> Use LevelDB For Backend
> -----------------------
>
> Key: ZOOKEEPER-3352
> URL: https://issues.apache.org/jira/browse/ZOOKEEPER-3352
> Project: ZooKeeper
> Issue Type: New Feature
> Components: server
> Reporter: David Mollitor
> Assignee: David Mollitor
> Priority: Critical
> Fix For: 4.0.0
>
>
> Use LevelDB for managing data stored in ZK (transaction logs and snapshots).
> https://stackoverflow.com/questions/6779669/does-leveldb-support-java
--
This message was sent by Atlassian JIRA
(v7.6.3#76005)
