Re: Is there a known bug with SKIP LOCKED and "tuple to be locked was already moved to another partition due to concurrent update"?

Hi everyone,

Is this a well know bug? I just hit the same issue. Below are steps to
reproduce.

Create data table and partition it by state

CREATE TABLE data (
id bigserial not null,
state smallint not null DEFAULT 1,
updated_at timestamp without time zone default now()
) partition by list(state);create table data_pending partition of data
for values in (1);create table data_processing partition of data for
values in (2);create table data_done partition of data for values in
(3);

Generate test data

INSERT INTO transactionsSELECT generate_series(1,1000) AS id, 1 AS state, NOW();

Move data from pending state to processing in batches

UPDATE transactionsSET transactions.state = 2, updated_at = NOW()WHERE id = (
SELECT id
FROM transactions
WHERE state = 1
LIMIT 10 FOR UPDATE SKIP LOCKED
)
RETURNING id;

You can now process them in application and move to done state when
finished.

However, this doesn't work as FOR UPDATE SKIP LOCKED fails for partitioned
tables with following error

ERROR: tuple to be locked was already moved to another partition due to
concurrent update

Here is full script to test this

cat > test-skip-locked-with-partitions << 'SCRIPT'PSQL_CMD="psql -q -U
postgres"eval $PSQL_CMD > /dev/null << EOFCREATE TABLE IF NOT EXISTS
data ( id bigserial not null, state smallint not null DEFAULT 1,
updated_at timestamp without time zone default now()) partition by
list(state);CREATE TABLE IF NOT EXISTS data_pending partition of data
for values in (1);CREATE TABLE IF NOT EXISTS data_processing partition
of data for values in (2);CREATE TABLE IF NOT EXISTS data_done
partition of data for values in (3);INSERT INTO dataSELECT
generate_series(1,1000) AS id, 1 AS state, NOW();EOFfunction run {eval
$PSQL_CMD > /dev/null << EOF UPDATE data SET state = 2,
updated_at = NOW() WHERE id IN ( SELECT id FROM data
WHERE state = 1 LIMIT 10 FOR UPDATE SKIP LOCKED )
RETURNING id;EOF}for i in {1..100}; do (run &) doneSCRIPT
bash test-skip-locked-with-partitions

Which results in

root(at)e92e3fa7fecf:/# bash test-skip-locked-with-partitions
ERROR: tuple to be locked was already moved to another partition due
to concurrent update
ERROR: tuple to be locked was already moved to another partition due
to concurrent update
ERROR: tuple to be locked was already moved to another partition due
to concurrent update
ERROR: tuple to be locked was already moved to another partition due
to concurrent update
ERROR: tuple to be locked was already moved to another partition due
to concurrent update
ERROR: tuple to be locked was already moved to another partition due
to concurrent update
ERROR: tuple to be locked was already moved to another partition due
to concurrent update
ERROR: tuple to be locked was already moved to another partition due
to concurrent update
ERROR: tuple to be locked was already moved to another partition due
to concurrent update

This works well if table is not partitioned. However, query with WHERE
state=1 has hard time when there are millions of records. Adding index
doesn't help as Query Planner tends to ignore it and do Seq Scan anyway.
Workaround for now is to create separate queue table, which is like custom
partitioning.

INSERT SAD CAT PICTURE

https://gist.github.com/arvenil/b46e927c943fa7495780ea2ae5492e78

Best, Kamil Dziedzic

On Thu, Mar 11, 2021 at 10:08 PM Jim Jarvie <jim(at)talentstack(dot)to> wrote:

> Hi Gunther & List,
>
> I think I have an extremely similar issue and things point in the same
> direction of a potential issue for skip locked on partitioned tables.
>
> Background is I had a queue table on v9.6 with fairly high volume (>50M
> items, growth in the 1+M/daily).
>
> Processing the queue with FOR UPDATE SKIP LOCKED was reliable but traffic
> volumes on v9.6 and the fact v12 is current let to migrating to v12 and
> using a partitioned table.
>
> Queue has distinct categories of items, so the table is partitioned by
> list on each category. Processing in 1 category results in it being
> updated to the next logical category (i.e. it will migrate partition once
> it is processed).
>
> Within each category, there can be 10'sM rows, so the list partition is
> hash partitioned as well. I don't think this is the issue but is mentioned
> for completeness.
>
> Now, when processing the queue, there are regular transaction aborts with
> "tuple to be locked was already moved to another partition due to
> concurrent update".
>
> From everything I can trace, it really does look like this is caused by
> rows which should be locked/skipped as they are processed by a different
> thread.
>
> I tried switching 'for update' to 'for key share' and that created a
> cascade of deadlock aborts, so was worse for my situation.
>
> For now, I roll back and repeat the select for update skip locked until it
> succeeds - which it eventually does.
>
> However, it really feels like these should just have been skipped by
> PostgreSQL without the rollback/retry until success.
>
> So, am I missing something/doing it wrong? Or could there be a potential
> issue that needs raised?
>
> Thanks
>
> Jim
>
>
>
> On 30-Jun.-2020 12:10, Gunther Schadow wrote:
>
> Hi all,
>
> long time ago I devised with your help a task queuing system which uses
> SELECT ... FOR UPDATE SKIP LOCKED for many parallel workers to find tasks
> in the queue, and it used a partitioned table where the hot part of the
> queue is short and so the query for a job is quick and the skip locked
> locking makes sure that one job is only assigned to one worker. And this
> works pretty well for me, except that when we run many workers we find a
> lot of these failures occurring:
>
> "tuple to be locked was already moved to another partition due to
> concurrent update"
>
> This would not exactly look like a bug, because the message says "to be
> locked", so at least it's not allowing two workers to lock the same tuple.
> But it seems that the skip-locked mode should not make an error out of
> this, but treat it as the tuple was already locked. Why would it want to
> lock the tuple (representing the job) if another worker has already
> finished his UPDATE of the job to mark it as "done" (which is what makes
> the tuple move to the "completed" partition.)
>
> Either the SELECT for jobs to do returned a wrong tuple, which was already
> update, or there is some lapse in the locking.
>
> Either way it would seem to be a waste of time throwing all these errors
> when the tuple should not even have been selected for update and locking.
>
> I wonder if anybody knows anything about that issue? Of course you'll want
> to see the DDL and SQL queries, etc. but you can't really try it out unless
> you do some massively parallel magic. So I figured I just ask.
>
> regards,
> -Gunther
>
>
>
>
>
>