Benjamin Reed commented on ZOOKEEPER-507:

here are the main changes:

1) changed to using netty for client requests
2) made recovery more reliable
3)  multiplexed writes to handle a large number of concurrent ledgers being 
written to

the user facing changes are minimal. our javadoc efforts have been with respect 
to user facing classes. the internal code is still a bit influx, so some of the 
internal or simple classes we have not put a lot of effort into documenting 
since they haven't really finalized.

our main goal right now is to get the patch in since it is very large and hard 
to manage at this point.

> BookKeeper client re-write
> --------------------------
>                 Key: ZOOKEEPER-507
>                 URL: https://issues.apache.org/jira/browse/ZOOKEEPER-507
>             Project: Zookeeper
>          Issue Type: Improvement
>          Components: contrib-bookkeeper
>            Reporter: Flavio Paiva Junqueira
>            Assignee: Utkarsh Srivastava
>             Fix For: 3.3.0
>         Attachments: BookieFailureTest-log-507.rtf, ZOOKEEPER-507.patch, 
> ZOOKEEPER-507.patch, ZOOKEEPER-507.patch, ZOOKEEPER-507.patch
> Error handling is far from ideal currently in the BookKeeper client.

This message is automatically generated by JIRA.
You can reply to this email to add a comment to the issue online.

Reply via email to