Re: [HACKERS] VACUUM FULL deadlock with backend startup

2011-03-22 Thread Tom Lane
Nikhil Sontakke nikhil.sonta...@enterprisedb.com writes: It looked familiar, so I dug up the archives and found that Tom had committed a fix for a similar deadlock via git commitid: 715120e7 However this current deadlock involved an index with oid 2663, which is ClassNameNspIndexId. Clearly

Re: [HACKERS] VACUUM FULL deadlock with backend startup

2011-03-22 Thread Nikhil Sontakke
Patch applied, thanks! Thanks Tom! Regards, Nikhils

Re: [HACKERS] VACUUM FULL deadlock with backend startup

2011-03-20 Thread Tom Lane
Robert Haas robertmh...@gmail.com writes: On Sat, Mar 19, 2011 at 10:46 AM, Nikhil Sontakke nikhil.sonta...@enterprisedb.com wrote: Not really a performance issue AFAICS. If the relcache init file exists, then the phase2 of the catalog cache which eventually calls the above code path is

Re: [HACKERS] VACUUM FULL deadlock with backend startup

2011-03-20 Thread Robert Haas
On Mar 20, 2011, at 2:24 AM, Tom Lane t...@sss.pgh.pa.us wrote: Robert Haas robertmh...@gmail.com writes: On Sat, Mar 19, 2011 at 10:46 AM, Nikhil Sontakke nikhil.sonta...@enterprisedb.com wrote: Not really a performance issue AFAICS. If the relcache init file exists, then the phase2 of the

Re: [HACKERS] VACUUM FULL deadlock with backend startup

2011-03-19 Thread Robert Haas
On Fri, Mar 18, 2011 at 11:25 PM, Nikhil Sontakke nikhil.sonta...@enterprisedb.com wrote: The fix is similar to the earlier commit by Tom. I tested this fix against 8.3.13. We lock the parent catalog now before calling index_open. Patch against git HEAD attached with this mail. I guess we will

Re: [HACKERS] VACUUM FULL deadlock with backend startup

2011-03-19 Thread Nikhil Sontakke
Hi, The fix is similar to the earlier commit by Tom. I tested this fix against 8.3.13. We lock the parent catalog now before calling index_open. Patch against git HEAD attached with this mail. I guess we will backpatch this? Tom's last commit was backpatched till 8.2 I think. Is it

Re: [HACKERS] VACUUM FULL deadlock with backend startup

2011-03-19 Thread Robert Haas
On Sat, Mar 19, 2011 at 10:46 AM, Nikhil Sontakke nikhil.sonta...@enterprisedb.com wrote: Not really a performance issue AFAICS. If the relcache init file exists, then the phase2 of the catalog cache which eventually calls the above code path is avoided. Oh, that doesn't sound so bad, then.

[HACKERS] VACUUM FULL deadlock with backend startup

2011-03-18 Thread Nikhil Sontakke
Hi, We encountered a deadlock involving VACUUM FULL (surprise surprise! :)) in PG 8.3.13 (and still not fixed in 9.0 AFAICS although the window appears much smaller). The call spits out the following deadlock info: ERROR: SQLSTATE 40P01: deadlock detected DETAIL: Process 12479 waits for

[HACKERS] VACUUM FULL deadlock with backend startup

2011-03-18 Thread Nikhil Sontakke
Hi, We encountered a deadlock involving VACUUM FULL (surprise surprise! :)) in PG 8.3.13 (and still not fixed in 9.0 AFAICS although the window appears much smaller). The call spits out the following deadlock info: ERROR: SQLSTATE 40P01: deadlock detected DETAIL: Process 12479 waits for