Обсуждение: Behaviour when autovacuum is canceled

On Thu, Sep 13, 2018 at 3:45 PM, Martín Fernández <fmartin91@gmail.com> wrote:

From what I could understand (that can be totally wrong), the vacuum process is split in multiple small transactions. If the autovacuum is canceled, could it be possible that only the latest transaction work be lost 

From the docs:

"VACUUM cannot be executed inside a transaction block."

As it is non-transactional any work it performs is live immediately and irrevocably as it occurs.

David J.

David,

Thanks a lot for the quick reply. 

I clearly misunderstood the references in the code. 

Best,

Martín

On Thu, Sep 13th, 2018 at 7:55 PM, "David G. Johnston" <david.g.johnston@gmail.com> wrote:

On Thu, Sep 13, 2018 at 3:45 PM, Martín Fernández <fmartin91@gmail.com> wrote:

From what I could understand (that can be totally wrong), the vacuum process is split in multiple small transactions. If the autovacuum is canceled, could it be possible that only the latest transaction work be lost 

From the docs:

"VACUUM cannot be executed inside a transaction block."

As it is non-transactional any work it performs is live immediately and irrevocably as it occurs.

David J.

Martín Fernández <fmartin91@gmail.com> writes:
> We basically started a VACUUM on a given table, waited for one index to process (captured cleaned rows count) and cancel the VACUUM. When we run another VACUUM on the same table the dead rows removed from the first index was a number slightly higher than the value logged on the first VACUUM. This behaviour made us feel that the work done to clean dead tuples on the first index was performed again. 

The unit of work that doesn't have to be repeated if VACUUM is canceled
is:

1. Scan a bunch of heap pages to identify dead tuples;
2. Scan *all* the table's indexes to remove the corresponding index entries;
3. Rescan those heap pages to actually remove the tuples.

It sounds like you canceled partway through phase 2.

The actual size of this unit of work is the number of dead-tuple TIDs
that will fit in maintenance_work_mem (at six or eight bytes apiece,
I forget whether it's aligned...). Normally, people make
maintenance_work_mem big so that they can reduce the number of index
scan cycles needed to complete vacuuming a table. But if you're
concerned about reducing the amount of work lost to a cancel,
you might try *reducing* maintenance_work_mem. This will make
vacuum slower overall (more index scans), but you have a better
chance that it will manage to actually remove some tuples before
getting canceled.

Or you could look at fixing the access patterns that are causing
so many autovacuum cancels.

regards, tom lane

On 2018-Sep-13, Martín Fernández wrote:

> By performing this changes we are going to start relying more heavily
> on the autovacuum work and the concern of "lost work" caused by
> autovacuum canceling itself when locking contention happen showed up.
> I'm guessing that we might be over thinking this and the canceling is
> not going to happen as frequently as we think it will.

Any DDL run on a table will cancel an autovacuum over that table (except
for-wraparound autovacuums). If these are rare, you don't need to worry
about that too much. If they are frequent enough that autovacuum will
be cancelled regularly in one table, you'll be sad.

If you're running vacuum by hand, you'd probably see your DDL blocking
behind VACUUM, which would be very noticeable. I think if you don't
have trouble today without having tuned the system carefully to avoid
such trouble, you're not likely to have trouble with autovacuum either.

--
Álvaro Herrera https://www.2ndQuadrant.com/
PostgreSQL Development, 24x7 Support, Remote DBA, Training & Services