First-draft release notes for next week's releases

classic Classic list List threaded Threaded
10 messages Options
Reply | Threaded
Open this post in threaded view
|

First-draft release notes for next week's releases

Tom Lane-2
... are committed at

https://git.postgresql.org/gitweb/?p=postgresql.git;a=commitdiff;h=5996cfc4665735a7e6e8d473bd66e8b11e320bbb

Please send comments/corrections by Sunday.

                        regards, tom lane

Reply | Threaded
Open this post in threaded view
|

Re: First-draft release notes for next week's releases

Peter Geoghegan-4
On Fri, Feb 8, 2019 at 5:19 PM Tom Lane <[hidden email]> wrote:
> Please send comments/corrections by Sunday.

Note that "Fix deadlock in GIN vacuum introduced by 218f51584d5"
(which is commit fd83c83d on the master branch) is not that far from
being a complete revert of a v10 feature (this appears as "Reduce page
locking during vacuuming of GIN indexes" in the v10 release notes).
Perhaps that's something that needs to be pointed out directly, as
happened with the the recheck_on_update issue in the last point
release. We may even need to revise the v10 release notes.

--
Peter Geoghegan

Reply | Threaded
Open this post in threaded view
|

Re: First-draft release notes for next week's releases

Tom Lane-2
Peter Geoghegan <[hidden email]> writes:
> Note that "Fix deadlock in GIN vacuum introduced by 218f51584d5"
> (which is commit fd83c83d on the master branch) is not that far from
> being a complete revert of a v10 feature (this appears as "Reduce page
> locking during vacuuming of GIN indexes" in the v10 release notes).

Yeah, I saw that in the commit message, but didn't really think that
the release note entry needed to explain it that way.  Could be
argued differently though.

> We may even need to revise the v10 release notes.

Perhaps just remove that item from the 10.0 notes?

                        regards, tom lane

Reply | Threaded
Open this post in threaded view
|

Re: First-draft release notes for next week's releases

Peter Geoghegan-4
On Fri, Feb 8, 2019 at 6:00 PM Tom Lane <[hidden email]> wrote:
> Yeah, I saw that in the commit message, but didn't really think that
> the release note entry needed to explain it that way.  Could be
> argued differently though.

I'm pretty confident that somebody is going to miss this
functionality, if this account of how the patch helped Yandex is
anything to go by:

https://www.postgresql.org/message-id/7B44397E-5E0A-462F-8148-1C444640FA0B%40simply.name

> > We may even need to revise the v10 release notes.
>
> Perhaps just remove that item from the 10.0 notes?

The wording could be changed to reflect the current reality within
GIN. It's still useful that posting trees are only locked when there
are pages to be deleted.

--
Peter Geoghegan

Reply | Threaded
Open this post in threaded view
|

Re: First-draft release notes for next week's releases

akapila
In reply to this post by Tom Lane-2
On Sat, Feb 9, 2019 at 6:49 AM Tom Lane <[hidden email]> wrote:
>
> ... are committed at
>
> https://git.postgresql.org/gitweb/?p=postgresql.git;a=commitdiff;h=5996cfc4665735a7e6e8d473bd66e8b11e320bbb
>

Author: Alvaro Herrera <[hidden email]>
+Branch: master [0464fdf07] 2019-01-21 20:08:52 -0300
+Branch: REL_11_STABLE [123cc697a] 2019-01-21 19:59:07 -0300
+-->
+     <para>
+      Create or delete foreign key enforcement triggers correctly when
+      attaching or detaching a partition in a a partitioned table that
+      has a foreign-key constraint (Amit Langote, &Aacute;lvaro Herrera)
+     </para>
+    </listitem>

It seems like there is a typo in the above sentence. /a a
partitioned/a partitioned

--
With Regards,
Amit Kapila.
EnterpriseDB: http://www.enterprisedb.com

Reply | Threaded
Open this post in threaded view
|

Re: First-draft release notes for next week's releases

Tom Lane-2
Amit Kapila <[hidden email]> writes:
> On Sat, Feb 9, 2019 at 6:49 AM Tom Lane <[hidden email]> wrote:
> +      Create or delete foreign key enforcement triggers correctly when
> +      attaching or detaching a partition in a a partitioned table that
> +      has a foreign-key constraint (Amit Langote, &Aacute;lvaro Herrera)

> It seems like there is a typo in the above sentence. /a a
> partitioned/a partitioned

Ooops, obviously my eyes had glazed over by the time I went back to
proofread.  Thanks for catching that.

                        regards, tom lane

Reply | Threaded
Open this post in threaded view
|

Re: First-draft release notes for next week's releases

Alexander Kuzmenkov
In reply to this post by Tom Lane-2
El 9/2/19 a las 4:19, Tom Lane escribió:
Please send comments/corrections by Sunday.

+      tuple deletion WAL record (Stas Kelvish)

-- a typo in his surname, should be Kelvich.

-- 
Alexander Kuzmenkov
Postgres Professional: http://www.postgrespro.com
The Russian Postgres Company
Reply | Threaded
Open this post in threaded view
|

Re: First-draft release notes for next week's releases

Tom Lane-2
In reply to this post by Peter Geoghegan-4
Peter Geoghegan <[hidden email]> writes:
> On Fri, Feb 8, 2019 at 6:00 PM Tom Lane <[hidden email]> wrote:
>> Yeah, I saw that in the commit message, but didn't really think that
>> the release note entry needed to explain it that way.  Could be
>> argued differently though.

> I'm pretty confident that somebody is going to miss this
> functionality, if this account of how the patch helped Yandex is
> anything to go by:
> https://www.postgresql.org/message-id/7B44397E-5E0A-462F-8148-1C444640FA0B%40simply.name

Ugh.  Well, hopefully somebody will find a less buggy solution
in the future.

>>> We may even need to revise the v10 release notes.

>> Perhaps just remove that item from the 10.0 notes?

> The wording could be changed to reflect the current reality within
> GIN. It's still useful that posting trees are only locked when there
> are pages to be deleted.

The v10 release notes just say

        Reduce page locking during vacuuming of <acronym>GIN</acronym> indexes
        (Andrey Borodin)

so it doesn't seem like there's any difference at that level of detail.
But I'll expand the new release note, say

      This change partially reverts a performance improvement, introduced
      in version 10, that attempted to reduce the number of index pages
      locked during deletion of a GIN posting tree page.  That's now been
      found to lead to deadlocks, so we've removed it pending closer
      analysis.

                        regards, tom lane

Reply | Threaded
Open this post in threaded view
|

Re: First-draft release notes for next week's releases

Peter Geoghegan-4
On Sun, Feb 10, 2019 at 10:05 AM Tom Lane <[hidden email]> wrote:

> Peter Geoghegan <[hidden email]> writes:
> > On Fri, Feb 8, 2019 at 6:00 PM Tom Lane <[hidden email]> wrote:
> The v10 release notes just say
>
>         Reduce page locking during vacuuming of <acronym>GIN</acronym> indexes
>         (Andrey Borodin)
>
> so it doesn't seem like there's any difference at that level of detail.
> But I'll expand the new release note, say
>
>       This change partially reverts a performance improvement, introduced
>       in version 10, that attempted to reduce the number of index pages
>       locked during deletion of a GIN posting tree page.  That's now been
>       found to lead to deadlocks, so we've removed it pending closer
>       analysis.

That plan seems sensible to me. The wording looks good, too.

--
Peter Geoghegan

Reply | Threaded
Open this post in threaded view
|

Re: First-draft release notes for next week's releases

Michael Paquier-2
In reply to this post by Alexander Kuzmenkov
On Sat, Feb 09, 2019 at 02:26:03PM +0300, Alexander Kuzmenkov wrote:
> El 9/2/19 a las 4:19, Tom Lane escribió:
> > Please send comments/corrections by Sunday.
>
> +      tuple deletion WAL record (Stas Kelvish)
>
> -- a typo in his surname, should be Kelvich.

You are right, the commit message is the origin of the mistake.  My
apologies.
--
Michael

signature.asc (849 bytes) Download Attachment