I have seen the same with all the *stat commands (iostat, vmstat, etc). You should ignore the output for the 1st interval.
-tom On Dec 26, 2007 12:32 PM, Imri Zvik <[EMAIL PROTECTED]> wrote: > I'm guessing that without an interval (or if with interval - the first > output), it is an average since boot. > > > > -----Original Message----- > > From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED] On Behalf Of > > Vitaly > > Sent: Wednesday, December 26, 2007 11:29 AM > > To: linux-il@cs.huji.ac.il > > Subject: Re: Interrups statistic - "sar" vs. "mstat" > > > > On Dec 26, 2007 10:04 AM, Vitaly Karasik <[EMAIL PROTECTED]> wrote: > > > Probaly it is trivial, but I don't understand why I see different > > statistics regarding interrups into "mstat" and "sar" output. There is > > 15997 against 92 !!! > > > Can someone explain it? > > > > > > [root]# sar -I SUM |head > > > Linux 2.4.21-47.ELsmp 12/25/2007 > > > > > > 12:00:01 AM INTR intr/s > > > 12:10:01 AM sum 15997.32 > > > > > > [ root]# mpstat > > > Linux 2.4.21-47.ELsmp 12/25/2007 > > > > > > 11:14:07 AM CPU %user %nice %system %iowait %irq %soft %idle intr/s > > > 11:14:07 AM all 2.40 0.00 7.41 0.05 1.97 10.78 77.39 92.36 > > > > > > > The answer was simple - "mpstat" provides wrong interup statistics > > when called without "interval" parameter. I mean, "mpstat 1" is OK, > > but "mpstat" will tell you wrong numbers. I don't understand yet, if > > this a feature or a bug. > > > > ================================================================= > > To unsubscribe, send mail to [EMAIL PROTECTED] with > > the word "unsubscribe" in the message body, e.g., run the command > > echo unsubscribe | mail [EMAIL PROTECTED] > > To unsubscribe, send mail to [EMAIL PROTECTED] with > the word "unsubscribe" in the message body, e.g., run the command > echo unsubscribe | mail [EMAIL PROTECTED] > > -- -tom 054-244-8025