You are viewing a plain text version of this content. The canonical link for it is here.

Posted to user@hive.apache.org by Greenhorn Techie <gr...@gmail.com> on 2018/05/18 11:24:57 UTC

HiveServer2 requests distribution

Hi,

We have our hive setup in a HA mode i.e. multiple HiveServer2 instances.
However, we believe the performance is not tuned properly and thinking
whether the requests are being skewed to only one or two hive servers
instead of being distributed across all the available HS2 instances.
Wondering if there is a mechanism to check how many requests have been
served by each HS2 instance. Any idea how to check this metric?

Thanks

Re: HiveServer2 requests distribution

Posted by Johannes Alberti <jo...@altiscale.com>.

You should be able to see this when enabling metrics, these are to my knowledge per instance metrics.

See also https://cwiki.apache.org/confluence/display/Hive/Hive+Metrics

Sent from my iPhone

> On May 21, 2018, at 2:53 AM, Greenhorn Techie <gr...@gmail.com> wrote:
> 
> Hi,
> 
> Any thoughts on this please?
> 
> Thanks
> 
> 
>> On 18 May 2018 at 12:24:57, Greenhorn Techie (greenhorntechie@gmail.com) wrote:
>> 
>> Hi,
>> 
>> We have our hive setup in a HA mode i.e. multiple HiveServer2 instances. However, we believe the performance is not tuned properly and thinking whether the requests are being skewed to only one or two hive servers instead of being distributed across all the available HS2 instances. Wondering if there is a mechanism to check how many requests have been served by each HS2 instance. Any idea how to check this metric? 
>> 
>> Thanks
>>

Re: HiveServer2 requests distribution

Posted by Greenhorn Techie <gr...@gmail.com>.

Hi,

Any thoughts on this please?

Thanks


On 18 May 2018 at 12:24:57, Greenhorn Techie (greenhorntechie@gmail.com)
wrote:

Hi,

We have our hive setup in a HA mode i.e. multiple HiveServer2 instances.
However, we believe the performance is not tuned properly and thinking
whether the requests are being skewed to only one or two hive servers
instead of being distributed across all the available HS2 instances.
Wondering if there is a mechanism to check how many requests have been
served by each HS2 instance. Any idea how to check this metric?

Thanks