Re: HELP on shared memory segments increasing in a MAD way

In article <wrbucmplma.fsf@tglobe3.tollpost.no>, ogj@tglobe3.tollpost.no says... > >Fernando.Collalti@orbit.net ("Collalti, Fernando") writes: > >> Hello, problems, problems, problems...... >> It seems that after less than 1 hour the engine collapses. >> With HELP from Neil Truby now is getting better(decreasing >> LRU_MAX_DIRTY and LRU_MIN_DIRTY) >> >> The system: HP9000 E35, 132MB RAM, Online 7.12.UC3 >> client/server application around 50 users >> was working fine with SE 5.04 and now....with OnLine... >> I'm taking Informix OnLine "off-line" each >> hour..... >> >> My parameters(continously changing and testing: >> LOCKS 2000> >Is 2000 locks enough? It depends on your setup, but I think it may be >too few. Monitor with 'onstat -p', look for 'ovlock'. > >> BUFFERS 40000(old=30000)> >I think this is too high for your memory configuration. What is the output >from 'onstat -g seg'? > >> NUMAIOVPS >> PHYSBUFF 32 >> LOGBUFF 32>> LOGSMAX 6 > >Do you have enough logs? 6 is pretty low. Monitor with 'onstat -l'. > >> CLEANERS 12(old=24) >> SHMBASE 0x0 >> SHMVIRTSIZE 60000 >> SHMADD 8192Make SHMTOTAL about half of size of physical memory. I would also lower some other parameters (e.g SHMVIRTSIZE 32768) >> SHMTOTAL 0 >> CKPTINTVL 300 >> LRUS 12(old=24)>> LRU_MAX_DIRTY 30(old=50) Good suggestion from Neil Truby before=70 >> LRU_MIN_DIRTY 20(old=40) before=60> >I have had good results (shorter checkpoint duration) with LRU_MAX_DIRTY=4 >and LRU_MIN_DIRTY=2 (yeah, I know. Pretty extreme values, but they were >in fact suggested by Informix Tech Support, and they helped a lot). > >> LTXHWM 50 >> LTXEHWM 60 >> TXTIMEOUT 0x12c >> STACKSIZE 32>> >> I've migrated from Informix-SE 5.04 to Informix-OnLine 7.12.UC3. >> With SE, the system was slow, but I never had problems. Now >> After the users are using the system(with the same client/server >> application) OnLine adds new shared memory segment (size 8388608) >> continuosly!!! >> After some hours(even one), the system "sits" down and I need to take OnLine >> OFF Line. I'm taking the OnLine 7.12 "on-line"with the NEW above values. >> I've change BUFFERS from 5000 to 30000(and now to 40000)and SHMVISTSIZE from >> >> 8000 >> to 60000. It seems that the application that was running OK on >> SE, now does not release memory until it doesn't disconnetcs >> from the Database engine....Why???? Are my new parameters ok? >> HELP!!!! > >I'm sure others know a lot more on these issues. >-- >'yvind Gjerstad Systems dept Tollpost-Globe AS N-6301 'ndalsnes/Norway >E-mail: ogj@it.tollpost.no Phone: +47 7122 6663 Fax: +47 7122 6694