hbase-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Lars George <lars.geo...@gmail.com>
Subject Re: HTable thread safety in 0.20.6
Date Fri, 04 Mar 2011 06:54:29 GMT
Hi Suraj.

AFAIK asynchbase talks wire protocol directly, i.e. it implements an
RPC endpoint and can therefore handle different remote cluster
versions by simply adjusting what it reads or writes.

Lars

On Fri, Mar 4, 2011 at 5:34 AM, Suraj Varma <svarma.ng@gmail.com> wrote:
> This is interesting - so, doesn't asynchbase have any dependency on hbase
> jar or hadoop jar? How does it achieve this version independence?
>
> Also - does this mean that we could just swap the zookeeper quorum to that
> of a different version cluster and achieve a no outage upgrade (from
> client's perspective)?
>
> Thanks,
> --Suraj
>
> On Sun, Feb 20, 2011 at 10:18 PM, tsuna <tsunanet@gmail.com> wrote:
>
>> On Fri, Feb 18, 2011 at 2:58 PM, Jean-Daniel Cryans <jdcryans@apache.org>
>> wrote:
>> > Like the javadoc says:
>> >
>> > This class is not thread safe. Use one instance per thread.
>>
>> Alternatively, you can look at asynchbase:
>> https://github.com/stumbleupon/asynchbase
>> It's designed to be thread-safe from the ground up and very efficient
>> in a multi-threaded application that uses HBase as a backend.  It also
>> works with all HBase versions, whereas with HTable you need to change
>> the .jar and restart your application whenever you want to switch
>> between different versions.
>>
>> --
>> Benoit "tsuna" Sigoure
>> Software Engineer @ www.StumbleUpon.com
>>
>

Mime
View raw message