SYSCS_UTIL.SYSCS_INPLACE_COMPRESS_TABLE versus SYSCS_UTIL.SYSCS_UPDATE_STATISTICS

[Alan Burlison]

Does SYSCS_UTIL.SYSCS_INPLACE_COMPRESS_TABLE recalculate statistics, or
is it necessary to do a SYSCS_UTIL.SYSCS_UPDATE_STATISTICS as well?  The
docs are unclear.

--
Alan Burlison
--

[Knut Anders Hatlen]

Alan Burlison <Alan.Burlison@...> writes:

> Does SYSCS_UTIL.SYSCS_INPLACE_COMPRESS_TABLE recalculate statistics,
> or is it necessary to do a SYSCS_UTIL.SYSCS_UPDATE_STATISTICS as well?
> The docs are unclear.

Hi Alan,

As far as I can see, SYSCS_INPLACE_COMPRESS_TABLE does not update the
index statistics. I couldn't find anything in the in-place compress code
that attempted to update the statistics, and it didn't appear to have
updated them when I monitored the CREATIONTIMESTAMP column in
SYS.SYSSTATISTICS.

I think much of the documentation about index statistics was written
before in-place compress was implemented, which is probably why the docs
are unclear. Could you please file a JIRA issue and point to the
sections that you found unclear? Thanks.

--
Knut Anders

[Alan Burlison]

Knut Anders Hatlen wrote:

> As far as I can see, SYSCS_INPLACE_COMPRESS_TABLE does not update the
> index statistics. I couldn't find anything in the in-place compress code
> that attempted to update the statistics, and it didn't appear to have
> updated them when I monitored the CREATIONTIMESTAMP column in
> SYS.SYSSTATISTICS.
>
> I think much of the documentation about index statistics was written
> before in-place compress was implemented, which is probably why the docs
> are unclear. Could you please file a JIRA issue and point to the
> sections that you found unclear? Thanks.

Hmm just noticed this in the docs for the related
SYSCS_UTIL.SYSCS_COMPRESS_TABLE procedure:

"The SYSCS_UTIL.SYSCS_COMPRESS_TABLE system procedure updates statistics
on all indexes as part of the index rebuilding process."

But no mention of that (either way) for
SYSCS_UTIL.SYSCS_INPLACE_COMPRESS_TABLE

--
Alan Burlison
--

[Alan Burlison]

Knut Anders Hatlen wrote:

> Could you please file a JIRA issue and point to the
> sections that you found unclear? Thanks.

https://issues.apache.org/jira/browse/DERBY-4289

--
Alan Burlison
--