Cpu vp usage

AIX 5.3 IDS 10.00.UC5 We have had reports of slow performance so I have been monitoring the system. When I check with onstat -g glo I see that the cpu vp usage is skewed towards the 1st cpu vp. In the latest example cpu vp 1 has used ~4000 cpu seconds, cpu vp 2 has used ~2000 cpu seconds, cpu vp 3 has used ~1000 seconds, cpu vp 4 has used ~300 cpu seconds, cpu vp 5 has used ~170 cpu seconds, and cpu cp 6 has only used ~60 cpu seconds. But when I monitor the system with AIX monitoring tools I am seeing long periods of time when one of the cpus is running at 100% utilisation. Should I be looking at the AIX tuning yet? Or is there something I can do at the IDS level first? Regards Malcolm

malcolm.iiug wrote: > AIX 5.3 > > IDS 10.00.UC5 > > > > We have had reports of slow performance so I have been monitoring the > system. When I check with onstat 'g glo I see that the cpu vp usage is > skewed towards the 1^st cpu vp. In the latest example cpu vp 1 has used > ~4000 cpu seconds, cpu vp 2 has used ~2000 cpu seconds, cpu vp 3 has > used ~1000 seconds, cpu vp 4 has used ~300 cpu seconds, cpu vp 5 has > used ~170 cpu seconds, and cpu cp 6 has only used ~60 cpu seconds. But > when I monitor the system with AIX monitoring tools I am seeing long > periods of time when one of the cpus is running at 100% utilisation. > > > > Should I be looking at the AIX tuning yet? Or is there something I can > do at the IDS level first? > > > > Regards > > > > Malcolm > > > > > Well, that sounds fine. BUT ... how many CPUs (or Cores) has the actual machine got? Perhaps you a doing some really big hash joins? When the CPU VP is running at 100% run a couple of onstat -g act's to see what is actually active.

On Aug 17, 4:31 am, "malcolm.iiug" <mali...@btopenworld.com> wrote: > AIX 5.3 > > IDS 10.00.UC5 > > We have had reports of slow performance so I have been monitoring the > system. When I check with onstat -g glo I see that the cpu vp usage is > skewed towards the 1st cpu vp. In the latest example cpu vp 1 has used > ~4000 cpu seconds, cpu vp 2 has used ~2000 cpu seconds, cpu vp 3 has used > ~1000 seconds, cpu vp 4 has used ~300 cpu seconds, cpu vp 5 has used ~170 > cpu seconds, and cpu cp 6 has only used ~60 cpu seconds. But when I monitor > the system with AIX monitoring tools I am seeing long periods of time when > one of the cpus is running at 100% utilisation. > > Should I be looking at the AIX tuning yet? Or is there something I can do at > the IDS level first? > > Regards > > Malcolm Two known causes: - If you are using shared memory connections in CPU VPs, you have to configure the same number of listeners as you have CPU VPs in the NETTYPE or only the first CPU VP is polling for work. Since the CPU VPs can only poll the shared memory locations (or TCP ports for that matter) when they are between tasks, the CPU VP that reads the input from the poll tends to pick up the requester's thread itself rather than placing it on the wait queue for another CPU VP. This results in the 1st CPU VP using about 70% of the total CPU time used by the CPU VPs and results in slower response time for new requests. - This one I just discovered when it hit us. With all current ESQL (and 4GL) releases, if the client is compiled with POSIX threads (also SOLaris threads on Sun) as opposed to non-threaded or threaded using DCE threads, and uses a TCP connection, for some reason only the first TCP listener ever does any work, there's no load balancing across the NET VPS (or presumably CPU VPs if the TCP polling's going on in CPU VPs). New bug reported this week, and IBM's looking into it. Art S. Kagel

Art S. Kagel wrote: > On Aug 17, 4:31 am, "malcolm.iiug" <mali...@btopenworld.com> wrote: >> AIX 5.3 >> >> IDS 10.00.UC5 >> >> We have had reports of slow performance so I have been monitoring the >> system. When I check with onstat -g glo I see that the cpu vp usage is >> skewed towards the 1st cpu vp. In the latest example cpu vp 1 has used >> ~4000 cpu seconds, cpu vp 2 has used ~2000 cpu seconds, cpu vp 3 has used >> ~1000 seconds, cpu vp 4 has used ~300 cpu seconds, cpu vp 5 has used ~170 >> cpu seconds, and cpu cp 6 has only used ~60 cpu seconds. But when I monitor >> the system with AIX monitoring tools I am seeing long periods of time when >> one of the cpus is running at 100% utilisation. >> >> Should I be looking at the AIX tuning yet? Or is there something I can do at >> the IDS level first? >> >> Regards >> >> Malcolm > > Two known causes: > > - If you are using shared memory connections in CPU VPs, you have to > configure the same number of listeners as you have CPU VPs in the > NETTYPE or only the first CPU VP is polling for work. Since the CPU > VPs can only poll the shared memory locations (or TCP ports for that > matter) when they are between tasks, the CPU VP that reads the input > from the poll tends to pick up the requester's thread itself rather > than placing it on the wait queue for another CPU VP. This results in > the 1st CPU VP using about 70% of the total CPU time used by the CPU > VPs and results in slower response time for new requests. > > - This one I just discovered when it hit us. With all current ESQL > (and 4GL) releases, if the client is compiled with POSIX threads (also > SOLaris threads on Sun) as opposed to non-threaded or threaded using > DCE threads, and uses a TCP connection, for some reason only the first > TCP listener ever does any work, there's no load balancing across the > NET VPS (or presumably CPU VPs if the TCP polling's going on in CPU > VPs). New bug reported this week, and IBM's looking into it. > > Art S. Kagel > > Can you elaborate a bit more on this last one? You say "listener"... do you mean the poll threads? The bug number or PMR id would be great... Just curious... it sounds very strange... at least to me. Regards. -- Fernando Nunes Portugal http://informix-technology.blogspot.com My email works... but I don't check it frequently...