Re: Out of memory error when doing an update with IN clause
От | Sean Shanny |
---|---|
Тема | Re: Out of memory error when doing an update with IN clause |
Дата | |
Msg-id | 3FF0FC0F.6010202@earthlink.net обсуждение исходный текст |
Ответ на | Re: Out of memory error when doing an update with IN clause (Tom Lane <tgl@sss.pgh.pa.us>) |
Список | pgsql-general |
Tom, We can give you access to the machine Tom. We need to know what sort of access you require. Since I don't have the ability to correspond with you via email due to the earthlink filter you have on could you send me another email address privately or a phone number or smoke signal so we can exchange the info you need? Thanks. --sean Tom Lane wrote: >Sean Shanny <shannyconsulting@earthlink.net> writes: > > >>I run this: >> >> > > > >>explain update f_commerce_impressions set servlet_key = 60 where >>servlet_key in (68,69,70,71,87,90,94,91,98,105,106); >>ERROR: out of memory >>DETAIL: Failed on request of size 1024. >> >> > >Well, I have to confess to total bafflement. AFAICS the overflowing >hash table must be the duplicate-tuple hash table that nodeIndexscan.c >sets up --- but that table shouldn't get any entries loaded into it >if you just do EXPLAIN with no ANALYZE. Furthermore, it should only >get loaded with entries for tuples that match the WHERE clause, and >you said earlier that there are no rows with these servlet_key values. >The code involved is all new in 7.4, so finding a bug in it wouldn't >surprise me much, but I can't see how this could be happening. > >It would help if you could rebuild with --enable-debug (if that wasn't >on already) and get a stack trace from the errfinish() call. Or, is >there any chance I could get access to your machine and look at the >problem for myself? > > regards, tom lane > >---------------------------(end of broadcast)--------------------------- >TIP 8: explain analyze is your friend > > >
В списке pgsql-general по дате отправления: