Re: BUG #16663: DROP INDEX did not free up disk space: idle connection hold file marked as deleted - Mailing list pgsql-hackers

From Nail Carpenter
Subject Re: BUG #16663: DROP INDEX did not free up disk space: idle connection hold file marked as deleted
Date
Msg-id CAA3qoJ=a_ksL5Y_oYTcpmysnwdW4r99KfhaeGbyyG23hX3F3FQ@mail.gmail.com
Whole thread Raw
In response to Re: BUG #16663: DROP INDEX did not free up disk space: idle connection hold file marked as deleted  (Pavel Borisov <pashkin.elfe@gmail.com>)
List pgsql-hackers
Yes, It's my fault. You're right.  

Pavel Borisov <pashkin.elfe@gmail.com> 于2020年11月19日周四 下午11:55写道:
One thing that doesn't matter is that the modify here seems unnecessary, right?

> mdunlinkfork(RelFileNodeBackend rnode, ForkNumber forkNum, bool isRedo)
> {
> char     *path;
> -     int                     ret;
> +     int                     ret = 0;
> path = relpath(rnode, forkNum

I suppose it is indeed necessary as otherwise the result of the comparison is not defined in case of 'else' block in the mdunlinkfork() :  
346     else
347     {
348         /* Prevent other backends' fds from holding on to the disk space */
349         do_truncate(path);
.....
356      * Delete any additional segments.
357      */
358     if (ret >= 0) 
----------^^^^^^^

--
Best regards,
Pavel Borisov

Postgres Professional: http://postgrespro.com
 
So in the present logic, *ret* is always 0 if it is not in recovery mode (and other *if* conditions are not satisfied). But when the *if* condition is satisfied, it is possible to skip the deletion of additional segments. 
Considering that our goal is to always try to unlink additional segments, *ret* seems unnecessary here. The code flow looks like:

> if (isRedo || .....)
> {
>     int ret;  /* move to here */
>     ....
> }
> else
> { }
>
> /* Delete any additional segments. */
> if (true)
> ...

Or is there any reason to allow us to skip the attempt to delete additional segments in recovery mode? 

pgsql-hackers by date:

Previous
From: Masahiko Sawada
Date:
Subject: Re: VACUUM (DISABLE_PAGE_SKIPPING on)
Next
From: James Hilliard
Date:
Subject: Re: [PATCH 1/1] Fix compilation on mac with Xcode >= 11.4.