From Bugzilla Helper: User-Agent: Mozilla/5.0 (X11; U; Linux i686; en-US; rv:0.9.7) Gecko/20011226 Description of problem: The code for reading extended statistics in iostat seems to be out of sync with the sard patch. iostat assumes that the time values in /proc/partitions are in ticks. This was true in the early versions of sard patch, but starting with sard-0.5, the values are converted to milliseconds. iostat performs a second "conversion," inflating the values by a factor of 10. This affects the columns avqu-sz, await, svctm, and %util. Version-Release number of selected component (if applicable): How reproducible: Always Steps to Reproduce: 1.iostat -x -d [ <device> ] 2. 3. Actual Results: Device: rrqm/s wrqm/s r/s w/s rsec/s wsec/s avgrq-sz avgqu-sz await svctm %util hdb 0.00 0.00 0.00 0.00 0.00 0.00 2.00 0.00 100.00 100.00 0.00 Expected Results: The corresponding line from /proc/partitions shows: major minor #blocks name rio rmerge rsect ruse wio wmerge wsect wuse running use aveq 3 64 6346368 hdb 2 0 4 20 0 0 0 0 0 20 20 The times are in milliseconds (known from kernel source) The disk had two read requests that were serviced in 20ms. The average wait time and service time should be 10.00, not 100.00. Additional info:
Is this still a problem with 4.0.3-1 from http://people.redhat.com/teg/sysstat/? (put there for other reasons, so I might as well find out :)
Ping?
Since Trond requested packages to be tested over 8 months ago, and got no response, and no response on a ping a month later, I'm closing this bug and assuming that it is fixed and working properly, either with 7.2 erratum (if any), or with a newer version of Red Hat Linux. If the problem still occurs in Red Hat Linux 7.2, it will not be addressed at this time, however if it occurs in Red Hat Linux 8.0 fully up2date, feel free to reopen the report and update it to point to 8.0. Assuming it is working in CURRENTRELEASE