maria-developers team mailing list archive

[Sergei Golubchik <serg@xxxxxxxxxxx>]

Hi, svoj!

On Jun 26, svoj@xxxxxxxxxxx wrote:
> revision-id: f37f19a6c790100cc2ae5f483ba4db34a517f0ae
> parent(s): 4a3f0684195703996b23c51af79cf946b85b8053
> committer: Sergey Vojtovich
> branch nick: 10.1
> timestamp: 2014-06-26 16:37:46 +0400
> message:
> 
> MDEV-4904 - Add query cache hit information to Percona Response time distribution plugin (MDEV-4568)
> 
> Added two new fields to INFORMATION_SCHEMA.QUERY_RESPONSE_TIME:
> QC_HITS_COUNT - query cache hits count
> QC_HITS_TOTAL - cumulative response time for query cache hits.

This is somewhat questionable feature. Queries served from qc don't take
much time, which means that @@query_response_time_range_base should be
different for them. Otherwise all qc hits will be in the first few
slots. But a different base means a different table
(QUERY_CACHE_RESPONSE_TIME ?) and a different set of configuration
variables.

On the other hand, this current implementation is much simpler.
So I think we can accept this patch. But I don't know how useful it'll be.
Alternatively, you can do a separate I_S table for QC.

As you like.

> diff --git a/plugin/query_response_time/mysql-test/query_response_time/query_response_time-stored.result b/plugin/query_response_time/mysql-test/query_response_time/query_response_time-stored.result
> index bec7007..e487fc5 100644
> --- a/plugin/query_response_time/mysql-test/query_response_time/query_response_time-stored.result
> +++ b/plugin/query_response_time/mysql-test/query_response_time/query_response_time-stored.result
> @@ -40,51 +40,51 @@ SHOW GLOBAL VARIABLES where Variable_name like 'QUERY_RESPONSE_TIME_RANGE_BASE';
>  Variable_name	Value
>  query_response_time_range_base	2
>  SELECT * FROM INFORMATION_SCHEMA.QUERY_RESPONSE_TIME;

I'd rather replaced * with TIME,COUNT,TOTAL - less changes and more
future proof.

> -TIME	COUNT	TOTAL
> -      0.000001	45	      0.000000
> +TIME	COUNT	TOTAL	QC_HITS_COUNT	QC_HITS_TOTAL
> +      0.000001	45	      0.000000	0	      0.000000
...
Regards,
Sergei