Thread: Re: logical decoding bug: segfault in ReorderBufferToastReplace()
On 12/13/19 16:13, Jeremy Schneider wrote:
On 12/11/19 08:35, Andres Freund wrote:done - in the commit that I replied to when I started this thread :)Seems like we clearly should add an elog(ERROR) here, so we error out, rather than crash.
https://git.postgresql.org/gitweb/?p=postgresql.git;a=commitdiff;h=69f883fef14a3fc5849126799278abcc43f40f56
Another PostgreSQL user ran into this issue. This time on version 12.5 - so instead of a crash they got the error message from the commit.
ERROR: XX000: could not open relation with OID 0
LOCATION: ReorderBufferToastReplace, reorderbuffer.c:305
Upon seeing this error message, I realized that the base relation OID would be very useful when the toast relation OID is "0".
Would this patch work to show that?
diff --git a/src/backend/replication/logical/reorderbuffer.c b/src/backend/replication/logical/reorderbuffer.c
index 2d9e1279bb..b90603b051 100644
--- a/src/backend/replication/logical/reorderbuffer.c
+++ b/src/backend/replication/logical/reorderbuffer.c
@@ -4598,8 +4598,8 @@ ReorderBufferToastReplace(ReorderBuffer *rb, ReorderBufferTXN *txn,
toast_rel = RelationIdGetRelation(relation->rd_rel->reltoastrelid);
if (!RelationIsValid(toast_rel))
- elog(ERROR, "could not open relation with OID %u",
- relation->rd_rel->reltoastrelid);
+ elog(ERROR, "could not open toast relation with OID %u (base relation with OID %u)",
+ relation->rd_rel->reltoastrelid, relation->rd_rel->oid);
toast_desc = RelationGetDescr(toast_rel);
Thoughts?
-Jeremy
-- Jeremy Schneider Database Engineer Amazon Web Services
On 2021-Jun-04, Jeremy Schneider wrote: > ERROR: XX000: could not open relation with OID 0 > LOCATION: ReorderBufferToastReplace, reorderbuffer.c:305 Hah. It seems to me that this code should silently return if rd_rel->reltoastrelid == 0; just like in the case of txn->toast_hash == NULL. It evidently means that no datum can be toasted, and therefor no toast replacement is needed. (As far as I recall, a table cannot go from having a toast table to not having one.) -- Álvaro Herrera Valdivia, Chile "¿Qué importan los años? Lo que realmente importa es comprobar que a fin de cuentas la mejor edad de la vida es estar vivo" (Mafalda)
On Sat, Jun 5, 2021 at 5:05 AM Alvaro Herrera <alvherre@alvh.no-ip.org> wrote: > > On 2021-Jun-04, Jeremy Schneider wrote: > > > ERROR: XX000: could not open relation with OID 0 > > LOCATION: ReorderBufferToastReplace, reorderbuffer.c:305 > > Hah. > > It seems to me that this code should silently return if > rd_rel->reltoastrelid == 0; just like in the case of > txn->toast_hash == NULL. It evidently means that no datum can be > toasted, and therefor no toast replacement is needed. > Even, if this fixes the issue, I guess it is better to find why this happens? I think the reason why the code is giving an error is that after toast insertions we always expect the insert on the main table of toast table, but if there happens to be a case where after toast insertion, instead of getting the insertion on the main table we get an insert in some other table then you will see this error. I think this can happen for speculative insertions where insertions lead to a toast table insert, then we get a speculative abort record, and then insertion on some other table. The main thing is currently decoding code ignores speculative aborts due to which such a problem can occur. Now, there could be other cases where such a problem can happen but if my theory is correct then the patch we are discussing in the thread [1] should solve this problem. Jeremy, is this problem reproducible? Can we get a testcase or pg_waldump output of previous WAL records? [1] - https://www.postgresql.org/message-id/CAExHW5sPKF-Oovx_qZe4p5oM6Dvof7_P%2BXgsNAViug15Fm99jA%40mail.gmail.com -- With Regards, Amit Kapila.
On 6/4/21 23:42, Amit Kapila wrote:
On 2021-Jun-04, Jeremy Schneider wrote:ERROR: XX000: could not open relation with OID 0 LOCATION: ReorderBufferToastReplace, reorderbuffer.c:305Even, if this fixes the issue, I guess it is better to find why this happens? I think the reason why the code is giving an error is that after toast insertions we always expect the insert on the main table of toast table, but if there happens to be a case where after toast insertion, instead of getting the insertion on the main table we get an insert in some other table then you will see this error. I think this can happen for speculative insertions where insertions lead to a toast table insert, then we get a speculative abort record, and then insertion on some other table. The main thing is currently decoding code ignores speculative aborts due to which such a problem can occur. Now, there could be other cases where such a problem can happen but if my theory is correct then the patch we are discussing in the thread [1] should solve this problem. Jeremy, is this problem reproducible? Can we get a testcase or pg_waldump output of previous WAL records? [1] - https://www.postgresql.org/message-id/CAExHW5sPKF-Oovx_qZe4p5oM6Dvof7_P%2BXgsNAViug15Fm99jA%40mail.gmail.com
It's unclear to me whether or not we'll be able to catch the repro on the actual production system. It seems that we are hitting this somewhat consistently, but at irregular and infrequent intervals. If we are able to catch it and walk the WAL records then I'll post back here. FYI, Bertrand was able to replicate the exact error message with pretty much the same repro that's in the other email thread which is linked above.
Separately, would there be any harm in adding the relation OID to the error message? Personally, I just think the error message is generally more useful if it shows the main relation OID (since we know that the toast OID can be 0). Not a big deal though.
-Jeremy
-- Jeremy Schneider Database Engineer Amazon Web Services
On Wed, Jun 9, 2021 at 12:06 AM Jeremy Schneider <schnjere@amazon.com> wrote: > > On 6/4/21 23:42, Amit Kapila wrote: > > On 2021-Jun-04, Jeremy Schneider wrote: > > ERROR: XX000: could not open relation with OID 0 > LOCATION: ReorderBufferToastReplace, reorderbuffer.c:305 > > Even, if this fixes the issue, I guess it is better to find why this > happens? I think the reason why the code is giving an error is that > after toast insertions we always expect the insert on the main table > of toast table, but if there happens to be a case where after toast > insertion, instead of getting the insertion on the main table we get > an insert in some other table then you will see this error. I think > this can happen for speculative insertions where insertions lead to a > toast table insert, then we get a speculative abort record, and then > insertion on some other table. The main thing is currently decoding > code ignores speculative aborts due to which such a problem can occur. > Now, there could be other cases where such a problem can happen but if > my theory is correct then the patch we are discussing in the thread > [1] should solve this problem. > > Jeremy, is this problem reproducible? Can we get a testcase or > pg_waldump output of previous WAL records? > > [1] - https://www.postgresql.org/message-id/CAExHW5sPKF-Oovx_qZe4p5oM6Dvof7_P%2BXgsNAViug15Fm99jA%40mail.gmail.com > > > It's unclear to me whether or not we'll be able to catch the repro on the actual production system. It seems that we arehitting this somewhat consistently, but at irregular and infrequent intervals. If we are able to catch it and walk theWAL records then I'll post back here. > Okay, one thing you can check is if there is a usage of Insert .. On Conflict .. statement in the actual production system? > FYI, Bertrand was able to replicate the exact error message with pretty much the same repro that's in the other email threadwhich is linked above. > > Separately, would there be any harm in adding the relation OID to the error message? Personally, I just think the errormessage is generally more useful if it shows the main relation OID (since we know that the toast OID can be 0). Nota big deal though. > I don't think that is a bad idea. However, I think it might be better to propose that as a separate patch in a new thread. -- With Regards, Amit Kapila.
Hi Amit, On 6/9/21 5:33 AM, Amit Kapila wrote: > On Wed, Jun 9, 2021 at 12:06 AM Jeremy Schneider <schnjere@amazon.com> wrote: >> On 6/4/21 23:42, Amit Kapila wrote: >> >> On 2021-Jun-04, Jeremy Schneider wrote: >> >> ERROR: XX000: could not open relation with OID 0 >> LOCATION: ReorderBufferToastReplace, reorderbuffer.c:305 >> >> Even, if this fixes the issue, I guess it is better to find why this >> happens? I think the reason why the code is giving an error is that >> after toast insertions we always expect the insert on the main table >> of toast table, but if there happens to be a case where after toast >> insertion, instead of getting the insertion on the main table we get >> an insert in some other table then you will see this error. I think >> this can happen for speculative insertions where insertions lead to a >> toast table insert, then we get a speculative abort record, and then >> insertion on some other table. The main thing is currently decoding >> code ignores speculative aborts due to which such a problem can occur. >> Now, there could be other cases where such a problem can happen but if >> my theory is correct then the patch we are discussing in the thread >> [1] should solve this problem. >> >> Jeremy, is this problem reproducible? Can we get a testcase or >> pg_waldump output of previous WAL records? >> >> [1] - https://www.postgresql.org/message-id/CAExHW5sPKF-Oovx_qZe4p5oM6Dvof7_P%2BXgsNAViug15Fm99jA%40mail.gmail.com >> >> >> It's unclear to me whether or not we'll be able to catch the repro on the actual production system. It seems that we arehitting this somewhat consistently, but at irregular and infrequent intervals. If we are able to catch it and walk theWAL records then I'll post back here. >> > Okay, one thing you can check is if there is a usage of Insert .. On > Conflict .. statement in the actual production system? Yes that's the case, so that a speculative abort record followed by an insert on some other table looks a perfect valid scenario regarding this current issue. Bertrand
On Wed, Jun 9, 2021 at 11:37 AM Drouvot, Bertrand <bdrouvot@amazon.com> wrote: > > On 6/9/21 5:33 AM, Amit Kapila wrote: > > On Wed, Jun 9, 2021 at 12:06 AM Jeremy Schneider <schnjere@amazon.com> wrote: > >> On 6/4/21 23:42, Amit Kapila wrote: > >> > >> On 2021-Jun-04, Jeremy Schneider wrote: > >> > >> ERROR: XX000: could not open relation with OID 0 > >> LOCATION: ReorderBufferToastReplace, reorderbuffer.c:305 > >> > >> Even, if this fixes the issue, I guess it is better to find why this > >> happens? I think the reason why the code is giving an error is that > >> after toast insertions we always expect the insert on the main table > >> of toast table, but if there happens to be a case where after toast > >> insertion, instead of getting the insertion on the main table we get > >> an insert in some other table then you will see this error. I think > >> this can happen for speculative insertions where insertions lead to a > >> toast table insert, then we get a speculative abort record, and then > >> insertion on some other table. The main thing is currently decoding > >> code ignores speculative aborts due to which such a problem can occur. > >> Now, there could be other cases where such a problem can happen but if > >> my theory is correct then the patch we are discussing in the thread > >> [1] should solve this problem. > >> > >> Jeremy, is this problem reproducible? Can we get a testcase or > >> pg_waldump output of previous WAL records? > >> > >> [1] - https://www.postgresql.org/message-id/CAExHW5sPKF-Oovx_qZe4p5oM6Dvof7_P%2BXgsNAViug15Fm99jA%40mail.gmail.com > >> > >> > >> It's unclear to me whether or not we'll be able to catch the repro on the actual production system. It seems that weare hitting this somewhat consistently, but at irregular and infrequent intervals. If we are able to catch it and walkthe WAL records then I'll post back here. > >> > > Okay, one thing you can check is if there is a usage of Insert .. On > > Conflict .. statement in the actual production system? > > Yes that's the case, so that a speculative abort record followed by an > insert on some other table looks a perfect valid scenario regarding this > current issue. > Okay, thanks for the confirmation. So the patch being discussed in that thread will fix your problem. -- With Regards, Amit Kapila.
Hi, On 6/9/21 8:10 AM, Amit Kapila wrote: > CAUTION: This email originated from outside of the organization. Do not click links or open attachments unless you canconfirm the sender and know the content is safe. > > > > On Wed, Jun 9, 2021 at 11:37 AM Drouvot, Bertrand <bdrouvot@amazon.com> wrote: >> On 6/9/21 5:33 AM, Amit Kapila wrote: >>> On Wed, Jun 9, 2021 at 12:06 AM Jeremy Schneider <schnjere@amazon.com> wrote: >>>> On 6/4/21 23:42, Amit Kapila wrote: >>>> >>>> On 2021-Jun-04, Jeremy Schneider wrote: >>>> >>>> ERROR: XX000: could not open relation with OID 0 >>>> LOCATION: ReorderBufferToastReplace, reorderbuffer.c:305 >>>> >>>> Even, if this fixes the issue, I guess it is better to find why this >>>> happens? I think the reason why the code is giving an error is that >>>> after toast insertions we always expect the insert on the main table >>>> of toast table, but if there happens to be a case where after toast >>>> insertion, instead of getting the insertion on the main table we get >>>> an insert in some other table then you will see this error. I think >>>> this can happen for speculative insertions where insertions lead to a >>>> toast table insert, then we get a speculative abort record, and then >>>> insertion on some other table. The main thing is currently decoding >>>> code ignores speculative aborts due to which such a problem can occur. >>>> Now, there could be other cases where such a problem can happen but if >>>> my theory is correct then the patch we are discussing in the thread >>>> [1] should solve this problem. >>>> >>>> Jeremy, is this problem reproducible? Can we get a testcase or >>>> pg_waldump output of previous WAL records? >>>> >>>> [1] - https://www.postgresql.org/message-id/CAExHW5sPKF-Oovx_qZe4p5oM6Dvof7_P%2BXgsNAViug15Fm99jA%40mail.gmail.com >>>> >>>> >>>> It's unclear to me whether or not we'll be able to catch the repro on the actual production system. It seems that weare hitting this somewhat consistently, but at irregular and infrequent intervals. If we are able to catch it and walkthe WAL records then I'll post back here. >>>> >>> Okay, one thing you can check is if there is a usage of Insert .. On >>> Conflict .. statement in the actual production system? >> Yes that's the case, so that a speculative abort record followed by an >> insert on some other table looks a perfect valid scenario regarding this >> current issue. >> > Okay, thanks for the confirmation. So the patch being discussed in > that thread will fix your problem. Yes, thanks a lot! Bertrand