Date: Fri, 17 Mar 2006 12:30:45 -0800 (PST) From: Dino Vliet <dino_vliet@yahoo.com> To: freebsd-questions@freebsd.org Subject: biord and sbwait state of postgresql process Message-ID: <20060317203045.45542.qmail@web51106.mail.yahoo.com>
next in thread | raw e-mail | index | archive | help
Hi folks, I'm running a application where I create multiple (300+) textfiles out of a postgresql database. The first 160 loops go perfectly but the suddenly the process of the app gets the state biord at certain point in the procedure. Then the WCPU column in the top command goes almost to 0% and also the CPU utilization drops to a few %. Normally both these values were showing values in the 80 and 90 range. The process stays here a few minutes (always at a certain point in the procedure) and runs very slowly but after all recovers. Then everything goes back to normal, and the process gets its state RUN again. The WCPU increases again to it's old values as does the CPU utilization but then my app goes back in another loop and starts degrading again. After I have cancelled the process, the postgres process still continues but now with a sbwait state. What does this mean? So, what is happening here? How can the performance degrade like that? And why always at the same step in the algorithm? Hope someone can help me out. Ooh yeah, I did some tweaks: sysctl -w kern.ipc.shmmax=402653184 sysctl -w kern.ipc.shmall=16384 I have an amd64 3000+ system with 1GB ram and a SATA harddisk running freebsd 6.0 release-p5. Thanks in advanced Dino __________________________________________________ Do You Yahoo!? Tired of spam? Yahoo! Mail has the best spam protection around http://mail.yahoo.com
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?20060317203045.45542.qmail>