Re: [HACKERS] ANALYZE locks pg_listener in EXCLUSIVE for long
At 02:54 PM 3/05/2004, Tom Lane wrote: Please dig deeper. I may have found the problem; all the hung processes show 'async_notify waiting' in ps, and the ANALYZE eventually dies with 'tuple concurrently updated'. The routine 'ProcessIncomingNotify' in async.c does indeed try to lock pg_listener (even if we're not using NOTIFY/LISTEN). Not sure why the ANALYZE is locking the relation, though...but it is locked in AccessShareLock. I can send a log of my investigations if necessary. Philip Warner| __---_ Albatross Consulting Pty. Ltd. |/ - \ (A.B.N. 75 008 659 498) | /(@) __---_ Tel: (+61) 0500 83 82 81 | _ \ Fax: (+61) 03 5330 3172 | ___ | Http://www.rhyme.com.au |/ \| |---- PGP key available upon request, | / and from pgp.mit.edu:11371 |/ ---(end of broadcast)--- TIP 9: the planner will ignore your desire to choose an index scan if your joining column's datatypes do not match
Re: [HACKERS] ANALYZE locks pg_listener in EXCLUSIVE for long
Further to this, ProcessIncomingNotify seems to hold the lock on the listener relation until it's current transaction exits. If the ANALYZE was not the source of the error, but was just another victim, does that mean it might hold the lock for a very long time if the analyze is lengthy? At 02:54 PM 3/05/2004, Tom Lane wrote: Please dig deeper. I may have found the problem; all the hung processes show 'async_notify waiting' in ps, and the ANALYZE eventually dies with 'tuple concurrently updated'. The routine 'ProcessIncomingNotify' in async.c does indeed try to lock pg_listener (even if we're not using NOTIFY/LISTEN). Not sure why the ANALYZE is locking the relation, though...but it is locked in AccessShareLock. I can send a log of my investigations if necessary. Philip Warner| __---_ Albatross Consulting Pty. Ltd. |/ - \ (A.B.N. 75 008 659 498) | /(@) __---_ Tel: (+61) 0500 83 82 81 | _ \ Fax: (+61) 03 5330 3172 | ___ | Http://www.rhyme.com.au |/ \| |---- PGP key available upon request, | / and from pgp.mit.edu:11371 |/ ---(end of broadcast)--- TIP 4: Don't 'kill -9' the postmaster
Re: [HACKERS] ANALYZE locks pg_listener in EXCLUSIVE for long
At 06:21 PM 3/05/2004, Philip Warner wrote: 'tuple concurrently updated' The database logs show the same error in each case where a long delay has occurred. And before anyone suggests it, we already have processes in place to prevent to ANALYZEs running at the same time. Philip Warner| __---_ Albatross Consulting Pty. Ltd. |/ - \ (A.B.N. 75 008 659 498) | /(@) __---_ Tel: (+61) 0500 83 82 81 | _ \ Fax: (+61) 03 5330 3172 | ___ | Http://www.rhyme.com.au |/ \| |---- PGP key available upon request, | / and from pgp.mit.edu:11371 |/ ---(end of broadcast)--- TIP 9: the planner will ignore your desire to choose an index scan if your joining column's datatypes do not match
Re: [HACKERS] ANALYZE locks pg_listener in EXCLUSIVE for long
At 06:21 PM 3/05/2004, Philip Warner wrote: 'tuple concurrently updated' I lied. The database DO NOT logs show the same error in each case where a long delay has occurred. It happens sometimes; recent process logs do show the 'async_notify waiting' status, however. I'll try not to send any more emails until someone responds ;-) Philip Warner| __---_ Albatross Consulting Pty. Ltd. |/ - \ (A.B.N. 75 008 659 498) | /(@) __---_ Tel: (+61) 0500 83 82 81 | _ \ Fax: (+61) 03 5330 3172 | ___ | Http://www.rhyme.com.au |/ \| |---- PGP key available upon request, | / and from pgp.mit.edu:11371 |/ ---(end of broadcast)--- TIP 5: Have you checked our extensive FAQ? http://www.postgresql.org/docs/faqs/FAQ.html
Re: [HACKERS] ANALYZE locks pg_listener in EXCLUSIVE for long time?
Philip Warner [EMAIL PROTECTED] writes: I may have found the problem; all the hung processes show 'async_notify waiting' in ps, and the ANALYZE eventually dies with 'tuple concurrently updated'. The routine 'ProcessIncomingNotify' in async.c does indeed try to lock pg_listener (even if we're not using NOTIFY/LISTEN). Not sure why the ANALYZE is locking the relation, though...but it is locked in AccessShareLock. Hm. What seems likely to have happened is that the sinval message queue got full. We currently deal with this by sending SIGUSR2 to all backends, which forces them through a NOTIFY-check cycle; a byproduct of the transaction start is to read pending sinval messages. (This is somebody's ugly quick hack from years ago; we really oughta find a less expensive way of doing it.) That would have left all the idle backends trying to get exclusive lock on pg_listener, and if the ANALYZE subsequently reached pg_listener, its share lock would queue up behind those requests. What is not clear yet is why *all* of them are blocked. Seems something else must have some kind of lock already on pg_listener; but who? Can you get a dump of the pg_locks view while this is happening? And before anyone suggests it, we already have processes in place to prevent to ANALYZEs running at the same time. How confident are you in those processes? I don't know of any other mechanism for 'tuple concurrently updated' failures in ANALYZE than concurrent analyze runs ... regards, tom lane ---(end of broadcast)--- TIP 7: don't forget to increase your free space map settings
Re: [HACKERS] ANALYZE locks pg_listener in EXCLUSIVE for long
At 07:33 PM 3/05/2004, Philip Warner wrote: I'll try not to send any more emails until someone responds ;-) I also noticed this in SIInsertDataEntry sinvaladt.c: /* * Try to prevent table overflow. When the table is 70% full send a * WAKEN_CHILDREN request to the postmaster. The postmaster will send * a SIGUSR2 signal (ordinarily a NOTIFY signal) to all the backends. * This will force idle backends to execute a transaction to look * through pg_listener for NOTIFY messages, and as a byproduct of the * transaction start they will read SI entries. * * This should never happen if all the backends are actively executing * queries, but if a backend is sitting idle then it won't be starting * transactions and so won't be reading SI entries. * * dz - 27 Jan 1998 */ Would a long-running ANALYZE (or other activity on a busy database) cause the shared buffers to get to the 70% threshold while doing a long-running ANALYZE? Philip Warner| __---_ Albatross Consulting Pty. Ltd. |/ - \ (A.B.N. 75 008 659 498) | /(@) __---_ Tel: (+61) 0500 83 82 81 | _ \ Fax: (+61) 03 5330 3172 | ___ | Http://www.rhyme.com.au |/ \| |---- PGP key available upon request, | / and from pgp.mit.edu:11371 |/ ---(end of broadcast)--- TIP 3: if posting/reading through Usenet, please send an appropriate subscribe-nomail command to [EMAIL PROTECTED] so that your message can get through to the mailing list cleanly
Re: [HACKERS] ANALYZE locks pg_listener in EXCLUSIVE for long
At 11:04 PM 3/05/2004, Tom Lane wrote: Hm. What seems likely to have happened is that the sinval message queue got full. I agree (our emails crossed). That would have left all the idle backends trying to get exclusive lock on pg_listener, and if the ANALYZE subsequently reached pg_listener, its share lock would queue up behind those requests. What I see is that the ANALYZE job already has it in ACCESS SHARED mode, and keeps the lock until it dies with the 'concurrent update' error. What is not clear yet is why *all* of them are blocked. Seems something else must have some kind of lock already on pg_listener; but who? ANALYZE. Can you get a dump of the pg_locks view while this is happening? Attached. How confident are you in those processes? I don't know of any other mechanism for 'tuple concurrently updated' failures in ANALYZE than concurrent analyze runs ... Fairly. In this particular instance the error was probably caused bu a manually run VACUUM (part of me stressing it to encourage the error). Contrary to my other email, we haven't had the 'tuple concurrently updated' error since March (until today, with me messing around). What I do have is minute-by-minute dumps of pg_locks and ps for the day. At each hang there were many processes in 'async_notify waiting' and an ANALYZE job had the lock in shared mode. I do not have minute-by-minute logs for more than today, but there were 3 hangs today, and only one with the concurrent update error. It would be interesting if we could find a piece of backend code that did a 'select * from pg_listener', and hence locked it in ACCESS SHARED. At the moment, it looks like either the ANALYZE is triggering an error that causes it's backend to read pg_listeners, or it is dying while ANALYZING pg_listeners. The latter seems unlikely since it hangs frequently, and pg_listeners is empty. Does ANALYZE rollback if it dies? Could this account for the delay? Philip Warner| __---_ Albatross Consulting Pty. Ltd. |/ - \ (A.B.N. 75 008 659 498) | /(@) __---_ Tel: (+61) 0500 83 82 81 | _ \ Fax: (+61) 03 5330 3172 | ___ | Http://www.rhyme.com.au |/ \| |---- PGP key available upon request, | / and from pgp.mit.edu:11371 |/ 18-02-hackers.dat.gz Description: Binary data ---(end of broadcast)--- TIP 5: Have you checked our extensive FAQ? http://www.postgresql.org/docs/faqs/FAQ.html
Re: [HACKERS] ANALYZE locks pg_listener in EXCLUSIVE for long time?
On Mon, May 03, 2004 at 02:14:18PM +1000, Gavin Sherry wrote: It is implemented using shared memory. I got stuck when I considered the situation where we rung out of shared memory. Some emails in the archive suggested we just fire all listeners but I didn't like that. Can this be kept in backend local memory and then sent to the other backends at transaction commit? If you run out of local memory you can just spill to disk. (With shared memory this seems pretty hard to do.) I'm not sure how would one send to the other backends. Maybe write another file on disk, one for each remote backend? Surely this can be done somehow. I've heard that on linux-2.6 they are implementing POSIX message queues (not sure what those are anyway); maybe we can do that on platforms that support it, for performance. -- Alvaro Herrera (alvherre[a]dcc.uchile.cl) In a specialized industrial society, it would be a disaster to have kids running around loose. (Paul Graham) ---(end of broadcast)--- TIP 9: the planner will ignore your desire to choose an index scan if your joining column's datatypes do not match
Re: [HACKERS] ANALYZE locks pg_listener in EXCLUSIVE for long time?
Philip Warner [EMAIL PROTECTED] writes: At 11:04 PM 3/05/2004, Tom Lane wrote: How confident are you in those processes? I don't know of any other mechanism for 'tuple concurrently updated' failures in ANALYZE than concurrent analyze runs ... Fairly. In this particular instance the error was probably caused bu a manually run VACUUM (part of me stressing it to encourage the error). Yeah, I see a process 14295 in your dump that seems to be trying to ANALYZE (at least, it's got write lock on pg_statistic...). 8631 is the incumbent ANALYZE, and it's got locks all over the place :-( I think what we have here is an evil side-effect of the change a couple versions back to allow standalone ANALYZE to run as a single transaction. A database-wide ANALYZE will therefore accumulate AccessShareLock on each table it touches, and it won't release these locks until commit. So the scenario goes like this: 1. Somewhere relatively early in its run, ANALYZE processes pg_statistic. So it's now holding AccessShareLock on pg_statistic. 2. As the ANALYZE proceeds, it issues sinval messages due to the updates it's making in pg_statistic. This is normal. There will be (at least) one such message per column analyzed, and it sounds like your database has many columns. Plus of course other catalog updates could be occurring in other backends. 3. There is at least one other backend sitting idle and therefore not reading sinval messages. So eventually the sinval queue gets 70% full and the SIGUSR2 escape-hatch is triggered. 4. The idle backends (and eventually non-idle ones, too, whenever they next reach the idle loop) try to do the NOTIFY thing, and get blocked trying to acquire AccessExclusiveLock on pg_listener. They will now be stuck until the ANALYZE completes. As a quick-hack fix, I think it would do to reduce the locks taken on pg_listener in async.c from AccessExclusiveLock to ExclusiveLock. This would serve the purpose of serializing async.c processing without creating a conflict against ANALYZE's AccessShareLock. Some other things we ought to think about for the future: * Is it really a good idea for database-wide ANALYZE to run as a single transaction? Holding all those locks is a recipe for deadlocks, even if they're as inoffensive as AccessShareLocks normally are. * Can we use something less heavyweight than ProcessIncomingNotify to deal with the sinval-clearing problem? Not only is that routine expensive, but it is a serialization bottleneck, which is exactly what we *don't* want in something that all the backends are getting told to do at the same time. I think the original motivation for that hack was because we didn't have a spare signal number available to dedicate to sinval response, but SIGUSR1 has been free for a couple releases now. I'm very tempted to commandeer it for sinval. regards, tom lane ---(end of broadcast)--- TIP 7: don't forget to increase your free space map settings
Re: [HACKERS] ANALYZE locks pg_listener in EXCLUSIVE for long time?
I'm not sure how would one send to the other backends. Maybe write another file on disk, one for each remote backend? Surely this can be done somehow. I've heard that on linux-2.6 they are implementing POSIX message queues (not sure what those are anyway); maybe we can do that on platforms that support it, for performance. Dunno if this is relevant, but if you want to go with message queues, there is also SystemV message queues. Since postgresql already uses sysv semaphores and shared memory, this would perhaps be portable to more systems that pg supports today (though you'd still need some kind of abstraction layer, since e.g. win32 does not have it). (man msgsnd, msgrcv, msgctl, msgget //Magnus ---(end of broadcast)--- TIP 9: the planner will ignore your desire to choose an index scan if your joining column's datatypes do not match
Re: [HACKERS] ANALYZE locks pg_listener in EXCLUSIVE for long time?
On Mon, 3 May 2004, Alvaro Herrera wrote: On Mon, May 03, 2004 at 02:14:18PM +1000, Gavin Sherry wrote: It is implemented using shared memory. I got stuck when I considered the situation where we rung out of shared memory. Some emails in the archive suggested we just fire all listeners but I didn't like that. Can this be kept in backend local memory and then sent to the other backends at transaction commit? If you run out of local memory you can just spill to disk. (With shared memory this seems pretty hard to do.) I'm not sure how would one send to the other backends. Maybe write another file on disk, one for each remote backend? Surely this can be done somehow. I've heard that on linux-2.6 they are implementing POSIX message queues (not sure what those are anyway); maybe we can do that on platforms that support it, for performance. What happens in the (unlikely) event that we never find space in shared memory? That's the problem that I am currently trying to solve. We currently just fire all the triggers but is that a great idea? Particularly if we support the passing of a message with a notify. Gavin ---(end of broadcast)--- TIP 8: explain analyze is your friend
Re: [HACKERS] ANALYZE locks pg_listener in EXCLUSIVE for long time?
Zeugswetter Andreas SB SD [EMAIL PROTECTED] writes: * Is it really a good idea for database-wide ANALYZE to run as a single transaction? Holding all those locks is a recipe for deadlocks, even if they're as inoffensive as AccessShareLocks normally are. Wasn't one idea behind that change also to not make the planner create a plan from mixed old and new statistics ? I don't recall that that was part of the discussion. IIRC all we were after was to let someone invoke ANALYZE from inside a BEGIN/COMMIT block. A possible compromise is to hack ANALYZE so that if it is invoked when *not* within a BEGIN block, it runs a separate transaction for each table. This seems pretty crufty but might satisfy all the requirements. I guess that could later be accomplished with begin work; analyze; commit work; (with subtransactions) though. AFAICS, locks taken by a (committed) subtransaction can't be released until top-level commit anyhow. Otherwise they fail to perform one of the essential functions of locking in an MVCC environment: to delay another process until the changes you've made are visible to him. regards, tom lane ---(end of broadcast)--- TIP 5: Have you checked our extensive FAQ? http://www.postgresql.org/docs/faqs/FAQ.html
Re: [HACKERS] ANALYZE locks pg_listener in EXCLUSIVE for long time?
I wrote: 2. As the ANALYZE proceeds, it issues sinval messages due to the updates it's making in pg_statistic. This is normal. Small correction: actually, backends only send sinval messages at commit, so the ANALYZE will just be accumulating pending messages in its private memory. Your observed symptom therefore can only occur if other transactions running parallel to the ANALYZE perform sufficient catalog updating activity to fill the sinval message queue. And there must also be at least one long-term-idle backend, so that the queue doesn't get drained. I had been wondering why we'd not identified this problem before, but that combination of factors is probably unusual enough to explain why not. regards, tom lane ---(end of broadcast)--- TIP 2: you can get off all lists at once with the unregister command (send unregister YourEmailAddressHere to [EMAIL PROTECTED])
Re: [HACKERS] ANALYZE locks pg_listener in EXCLUSIVE for long time?
* Is it really a good idea for database-wide ANALYZE to run as a single transaction? Holding all those locks is a recipe for deadlocks, even if they're as inoffensive as AccessShareLocks normally are. Wasn't one idea behind that change also to not make the planner create a plan from mixed old and new statistics ? I guess that could later be accomplished with begin work; analyze; commit work; (with subtransactions) though. Andreas ---(end of broadcast)--- TIP 4: Don't 'kill -9' the postmaster
Re: [HACKERS] ANALYZE locks pg_listener in EXCLUSIVE for long
At 01:30 AM 4/05/2004, Tom Lane wrote: can only occur if other transactions running parallel to the ANALYZE perform sufficient catalog updating activity to fill the sinval message queue. And there must also be at least one long-term-idle backend, so that the queue doesn't get drained. Sounds quite likely; usually seems to occur at 'shoulder' load times; lots of updates still happening (several each second) and a server process pool that is larger than necessary to handle the load. I'll replace all: heap_openr(ListenerRelationName, AccessExclusiveLock); with heap_openr(ListenerRelationName, ExclusiveLock); and see how it goes. Thanks for the help. Philip Warner| __---_ Albatross Consulting Pty. Ltd. |/ - \ (A.B.N. 75 008 659 498) | /(@) __---_ Tel: (+61) 0500 83 82 81 | _ \ Fax: (+61) 03 5330 3172 | ___ | Http://www.rhyme.com.au |/ \| |---- PGP key available upon request, | / and from pgp.mit.edu:11371 |/ ---(end of broadcast)--- TIP 7: don't forget to increase your free space map settings
Re: [HACKERS] ANALYZE locks pg_listener in EXCLUSIVE for long time?
Philip Warner [EMAIL PROTECTED] writes: Does this mean that ANALYZE will take an exclusive lock on pg_listener until the ANALYZE finishes? Or is there some other cause? ANALYZE does not take an exclusive lock on anything. However, the async.c functions want AccessExclusiveLock on pg_listener, so they quite possibly would get blocked by ANALYZE's not-so-exclusive lock. Possibly we could reduce the strength of the lock taken by the async.c functions ... I haven't thought hard about it. The long-term answer is certainly a wholesale rewrite of the listen/notify mechanism. regards, tom lane ---(end of broadcast)--- TIP 4: Don't 'kill -9' the postmaster
Re: [HACKERS] ANALYZE locks pg_listener in EXCLUSIVE for long time?
Tom Lane wrote: Philip Warner [EMAIL PROTECTED] writes: Does this mean that ANALYZE will take an exclusive lock on pg_listener until the ANALYZE finishes? Or is there some other cause? ANALYZE does not take an exclusive lock on anything. However, the async.c functions want AccessExclusiveLock on pg_listener, so they quite possibly would get blocked by ANALYZE's not-so-exclusive lock. Possibly we could reduce the strength of the lock taken by the async.c functions ... I haven't thought hard about it. The long-term answer is certainly a wholesale rewrite of the listen/notify mechanism. Gavin was working on it a while ago but I am not sure how far he got. -- Bruce Momjian| http://candle.pha.pa.us [EMAIL PROTECTED] | (610) 359-1001 + If your life is a hard drive, | 13 Roberts Road + Christ can be your backup.| Newtown Square, Pennsylvania 19073 ---(end of broadcast)--- TIP 1: subscribe and unsubscribe commands go to [EMAIL PROTECTED]
Re: [HACKERS] ANALYZE locks pg_listener in EXCLUSIVE for long
At 12:45 PM 3/05/2004, Tom Lane wrote: Possibly we could reduce the strength of the lock taken by the async.c functions If possible, this seems like a great option. We currently have a large database with several hundred users who get locked out for as much as half an hour while ANALYZE runs. The data in the database is extremely dynamic, so the analyze needs to be run regularly; we could run less often but this will just mean the problem happens once per week instead of once per day. Would ACCESS SHARE be OK? Philip Warner| __---_ Albatross Consulting Pty. Ltd. |/ - \ (A.B.N. 75 008 659 498) | /(@) __---_ Tel: (+61) 0500 83 82 81 | _ \ Fax: (+61) 03 5330 3172 | ___ | Http://www.rhyme.com.au |/ \| |---- PGP key available upon request, | / and from pgp.mit.edu:11371 |/ ---(end of broadcast)--- TIP 1: subscribe and unsubscribe commands go to [EMAIL PROTECTED]
Re: [HACKERS] ANALYZE locks pg_listener in EXCLUSIVE for long time?
Philip Warner [EMAIL PROTECTED] writes: Would ACCESS SHARE be OK? Certainly not, since the point of the locks in async.c is that only one backend should execute those routines at a time. ExclusiveLock might work okay ... but I still haven't thought hard about it ... regards, tom lane ---(end of broadcast)--- TIP 5: Have you checked our extensive FAQ? http://www.postgresql.org/docs/faqs/FAQ.html
Re: [HACKERS] ANALYZE locks pg_listener in EXCLUSIVE for long time?
Philip Warner [EMAIL PROTECTED] writes: If possible, this seems like a great option. We currently have a large database with several hundred users who get locked out for as much as half an hour while ANALYZE runs. If it takes half an hour to ANALYZE pg_listener, I think that ANALYZE is not your real problem :-(. You need a much more aggressive vacuuming policy on that table. Maybe a cron job issuing vacuum pg_listener once a minute would do? And get the size of the table knocked down to something less stratospheric to begin with --- perhaps stop all the listeners while you TRUNCATE the table. The existing listen/notify infrastructure isn't really designed for notification rates exceeding a few events per minute ... regards, tom lane ---(end of broadcast)--- TIP 6: Have you searched our list archives? http://archives.postgresql.org
Re: [HACKERS] ANALYZE locks pg_listener in EXCLUSIVE for long time?
On Sun, 2 May 2004, Bruce Momjian wrote: Tom Lane wrote: Philip Warner [EMAIL PROTECTED] writes: Does this mean that ANALYZE will take an exclusive lock on pg_listener until the ANALYZE finishes? Or is there some other cause? ANALYZE does not take an exclusive lock on anything. However, the async.c functions want AccessExclusiveLock on pg_listener, so they quite possibly would get blocked by ANALYZE's not-so-exclusive lock. Possibly we could reduce the strength of the lock taken by the async.c functions ... I haven't thought hard about it. The long-term answer is certainly a wholesale rewrite of the listen/notify mechanism. Gavin was working on it a while ago but I am not sure how far he got. Its basically written. It is implemented using shared memory. I got stuck when I considered the situation where we rung out of shared memory. Some emails in the archive suggested we just fire all listeners but I didn't like that. What I was considering was that when someone issues a NOTIFY, we reserve a slot for the NOTIFY (plus a message, which is why I originally looked at the problem) in shared memory. At the end of the transaction, we update a flag to say that the transaction successed or we remove it if we've aborted. Does anyone else have any thoughts about it? Gavin ---(end of broadcast)--- TIP 6: Have you searched our list archives? http://archives.postgresql.org
Re: [HACKERS] ANALYZE locks pg_listener in EXCLUSIVE for long
At 01:46 PM 3/05/2004, Tom Lane wrote: If it takes half an hour to ANALYZE pg_listener, I think that ANALYZE is not your real problem :-(. You need a much more aggressive vacuuming policy on that table. Maybe a cron job issuing vacuum pg_listener once a minute would do? And get the size of the table knocked down to something less stratospheric to begin with --- perhaps stop all the listeners while you TRUNCATE the table. It's a general ANALYZE command for the entire DB. It's about 6GB in size, and is vacuumed as frequently as possible; there is certainly unreclaimed space, but it does not substantially outweigh used space. My *guess* is that the largest table is being ANALYZEd at the time (it uses most of the 6GB), and for some reason pg_listeners is being locked in ACCESS SHARE the entire time. Just vacuuming pg_listener produces: vacuum verbose pg_listener; INFO: vacuuming pg_catalog.pg_listener INFO: pg_listener: found 0 removable, 0 nonremovable row versions in 0 pages VACUUM Philip Warner| __---_ Albatross Consulting Pty. Ltd. |/ - \ (A.B.N. 75 008 659 498) | /(@) __---_ Tel: (+61) 0500 83 82 81 | _ \ Fax: (+61) 03 5330 3172 | ___ | Http://www.rhyme.com.au |/ \| |---- PGP key available upon request, | / and from pgp.mit.edu:11371 |/ ---(end of broadcast)--- TIP 1: subscribe and unsubscribe commands go to [EMAIL PROTECTED]
Re: [HACKERS] ANALYZE locks pg_listener in EXCLUSIVE for long time?
Philip Warner [EMAIL PROTECTED] writes: ... for some reason pg_listeners is being locked in ACCESS SHARE the entire time. Just vacuuming pg_listener produces: vacuum verbose pg_listener; INFO: vacuuming pg_catalog.pg_listener INFO: pg_listener: found 0 removable, 0 nonremovable row versions in 0 pages VACUUM [blinks...] There's something pretty strange about that. Are you using LISTEN/NOTIFY at all? regards, tom lane ---(end of broadcast)--- TIP 9: the planner will ignore your desire to choose an index scan if your joining column's datatypes do not match
Re: [HACKERS] ANALYZE locks pg_listener in EXCLUSIVE for long
At 02:21 PM 3/05/2004, Tom Lane wrote: [blinks...] There's something pretty strange about that. Are you using LISTEN/NOTIFY at all? Nope. Philip Warner| __---_ Albatross Consulting Pty. Ltd. |/ - \ (A.B.N. 75 008 659 498) | /(@) __---_ Tel: (+61) 0500 83 82 81 | _ \ Fax: (+61) 03 5330 3172 | ___ | Http://www.rhyme.com.au |/ \| |---- PGP key available upon request, | / and from pgp.mit.edu:11371 |/ ---(end of broadcast)--- TIP 8: explain analyze is your friend
Re: [HACKERS] ANALYZE locks pg_listener in EXCLUSIVE for long time?
Philip Warner [EMAIL PROTECTED] writes: At 02:21 PM 3/05/2004, Tom Lane wrote: [blinks...] There's something pretty strange about that. Are you using LISTEN/NOTIFY at all? Nope. In that case there's no reason for anything to be taking any particular locks on pg_listener; and it's simply not possible for ANALYZE to spend half an hour on a zero-page table if it's not blocked by a lock. Could you dig a little deeper and see where the problem really is? regards, tom lane ---(end of broadcast)--- TIP 8: explain analyze is your friend
Re: [HACKERS] ANALYZE locks pg_listener in EXCLUSIVE for long
At 02:54 PM 3/05/2004, Tom Lane wrote: Please dig deeper. I will log everything I check next time; unfortunately, when it happens, the priority is on unlocking everything so I have a limited time to play. So far, killing the ANALYZE has fixed the problem each time. Philip Warner| __---_ Albatross Consulting Pty. Ltd. |/ - \ (A.B.N. 75 008 659 498) | /(@) __---_ Tel: (+61) 0500 83 82 81 | _ \ Fax: (+61) 03 5330 3172 | ___ | Http://www.rhyme.com.au |/ \| |---- PGP key available upon request, | / and from pgp.mit.edu:11371 |/ ---(end of broadcast)--- TIP 4: Don't 'kill -9' the postmaster
