Re: Logical Logs question

Steve Brown (steveb@sr.hp.com) wrote: : We are running ONLINE 5.02 UC1 on a HP9000 series 867 under HPUX 9.04. : Our tbconfig said LOGFILES 6 : LOGSIZE 500 : LOGSMAX 6 : These logs are defined in the root dbspace. : : We have been running fine for about a month with 1 database in one dbspace : and 2 databases in another dbspace, plus the root dbspace. : : We were testing the following sql code when we returned the error 458 - : Long transaction aborted. : update table1 : set div_status = (select div_stat_code from table2 : where table2.part_number = table1.part_number); : table1 has 115,000 records and table2 has 101,000 records. : By using the tbstat -l command I could see that all 6 logical logs were : filling up before the statement could complete. I know we can not change : the LOGSIZE without reloading all the data (over 900 MB), so I added : 4 more logs. The statement went farther, but still aborted. I added 5 : more logs for a total of 15 and it went to completion. : My question is: Is there an optimum balance between LOGFILES and LOGSIZE? : What is the maximum number of LOGFILES I can allocate? Steve, You can go up to about 60 logfiles. I prefer to make mine a lot bigger than yours. I usually go with 6 at about 25-30 meg. If your logs are not sufficient, you can get around it by breaking up your update job into several smaller jobs. Joe : Thanks for any input, : Steve Brown : Hewlett Packard : Santa Rosa, CA : email ... steveb@nmitd.sr.hp.com -- =========================================================================== jlumbley@netcom.com (Joe Lumbley) BancTec Service Corportation 214-450-9894 Dallas, Texas Watch for my _INFORMIX DBA SURVIVAL GUIDE_ in Fall '94 from Prentice Hall! ===========================================================================