> This paragraph reads a bit strange to me:
>
> + A statistics session is the time period when statistics are gathered by
> statistics collector
> + without being reset. So a statistics session continues across normal
> shutdowns,
> + but whenever statistics are reset, like during a crash or upgrade, a new
> time period
> + of statistics collection commences i.e. a new statistics session.
> + The query_id value generation is linked to statistics session to
> emphasize the fact
> + that whenever statistics are reset,the query_id for the same queries will
> also change.
>
> "time period when"? Shouldn't that be "time period during which".
> Also, doesn't a new "statistics session" start when a stats reset is
> invoked by the user? The bit after "commences" appears correct (to me,
> not a native by any means) but seems also a bit strange.
>
I have tried to rephrase this. Hopefully less confusing
A statistics session refers to the time period when statement
statistics are gathered by
statistics collector. A statistics session persists across normal
shutdowns. Whenever statistics are reset like during a crash or upgrade, a new
statistics session starts. The query_id value generation is linked to
statistics session to
emphasize that whenever statistics are reset,the query_id for the same
queries will also change.
regards
Sameer
View this message in context:
Re: pg_stat_statements: calls under-estimation propagation Sent from the
PostgreSQL - hackers mailing list archive at Nabble.com.
