Re: Syslocks table

I can imagine the lock problem, try this one for size: IDS 7.31.UC4 - AIX 4.3.3 A simple query with a union clause into a tempory table: (its inside a begin commit work) select count (*) cnt from table1 union select count (*) cnt from table2 into temp table3;select sum (cnt) from table3; drop table table3; Do this 60k odd times. What we experienced was : Lock usage increased quickly, 60k locks got wiped out in minutes, space usage in the dbspace where the temp table was got cleaned up in minutes again. Now that I think of it, I can possibly attribute it to the transaction. but lemme tell you, it was not amusing. This is what happens when you are in a hurry. brett In article <967019670.7195.0.nnrp-03.d4e5c55a@news.demon.co.uk>, pankaj@punks.demon.co.uk says... > Hi Art , > > How did you come to that conclusion ? > > Just want to know if my posting is wrong in my understanding of the way Informix > locking works. > > Cheers > > Pankaj > > Art S. Kagel wrote in message <39A2DDC3.74A54842@bloomberg.net>... > |This is a business critical problem. Tell the tech support engineer you > |want the case elevated to a priority level 2 (3 is default, 4 non-critical, > |1 the engineer does not get to go home until your problem is solved or he > |can hand off to a higher level engineer who is then in the same boat). > | > |Art S. Kagel > | > |Brian Fink wrote: > |> > |> Hello, > |> > |> We just upgraded to 7.31FC4 on HPUX. Everything seems to be working > |> fine except for a few lock overflows when trying to load a table. We > |> currently have 50,000 locks on the system and are trying to load in > |> about 30,000 records. We are averaging about 30 - 50 locks during > |> normal use. The settings are the same as the previous version since we > |> didn't need to do much tweaking (yet). > |> While researching the overflow problem I found that the syslocks table > |> doesn't contain any entries in it. When checking the status it shows a > |> null value for Number of Rows. Of course our 'onlocks' utility is a > |> bust since it reads from this table; however the onstat -k seems to > |> show the correct values. I say correct but I don't have any way to > |> check it. I have seen a few posts about the syslocks issue but nothing > |> referring to lock overflows. I've RTFM concerning performance and > |> upgrading in case of something different with the 64-bit but I can't > |> find an answer. My next step is support but I know that can take awhile. > |> > |> Any ideas? > |> > |> TIA > |> > |> Brian > |> > |> Sent via Deja.com http://www.deja.com/ > |> Before you buy. > > >