On Tue, Nov 13, 2018 at 11:32 AM Haribabu Kommi
<kommi.haribabu@gmail.com> wrote:
>
> On Mon, Nov 12, 2018 at 6:34 PM Amit Kapila <amit.kapila16@gmail.com> wrote:
>> > I can revert it back to void,
>> >
>>
>> +1, as we don't see any good reason to break backward compatibility.
>
>
> Thanks for the review.
> Attached the updated patch with return type as void.
>
With this patch, we are intending to remove the entries matching
userid, dbid, queryid from hash table (pgss_hash), but the contents of
the file (
pgss_query_texts.stat) will remain unchanged unless all of the entries
are removed from hash table. This appears fine to me, especially
because there is no easy way to remove the contents from the file.
Does anybody see any problem with this behavior?
Adding more info to the above point, even if the file contents are not
removed, later if the file size and number of pg_stat_statements entries
satisfy the garbage collection, the file will be truncated. So I feel not
removing of the contents when the query stats are reset using specific
parameters is fine.
Regards,
Haribabu Kommi
Fujitsu Australia