Probably https://issues.apache.org/jira/browse/HBASE-15654

Vote for it and ask developer to backport it to 1.1.x.

-Vlad

On Thu, Apr 28, 2016 at 6:14 PM, Michal Medvecky <[email protected]> wrote:

> Yes, it does.
>
> On Thu, Apr 28, 2016 at 5:58 PM, Vladimir Rodionov <[email protected]
> > wrote:
>
>> >> 2. Is it okay that one regionserver is 10x as loaded as others?
>>
>> Does this server host hbase:meta?
>>
>> -Vlad
>>
>> On Thu, Apr 28, 2016 at 4:14 PM, Michal Medvecky <[email protected]>
>> wrote:
>>
>>> Hello,
>>>
>>> I have 7-node hbase cluster (6 regionservers/hdfs nodes, 1 master), with
>>> one big table containing 500M+ entries.
>>>
>>> It's a heavy-write light-read database.
>>>
>>> 1. Is it okay that the table has 1000+ regions for 6 regionservers only?
>>>
>>> 2. Is it okay that one regionserver is 10x as loaded as others? See
>>> http://michal.medvecky.net/rs.png . This is a constant state, not
>>> random.
>>> Row keys are random, not sequential.
>>>
>>> 3. Is it okay that index table has 1000+ regions as well?
>>>
>>> The problem is that I need to speed up writes and it does not seem that
>>> adding node helps.
>>>
>>> I'm using hbase 1.1.3 + phoenix 4.6.0 on hadoop 2.7.1 on Ubuntu 14.04.
>>>
>>> Thank you
>>>
>>> Michal
>>>
>>
>>
>

Reply via email to