[RHEL 5.3] iostat data appears to be wrong
Issue
- iostat data for the dm devices does not add up to the total of the real usage on the devices that make up the dm
- iostat data has 'await' times less than 'service time', but await time includes service time according the iostat man page
- sample #1: dm device data doesn't correlate to individual underlying disks
- Based on the physical sd's if you add up their avgqu-sz from the same period it doesn't calculate.
- Expected results: sd's (1.28) would roll up into the dm (58.64).
- If look at dm-18 it has a very large avgqu-sz.
Device: rrqm/s wrqm/s r/s w/s rsec/s wsec/s avgrq-sz avgqu-sz await svctm %util
sda 0.00 246.40 0.00 6.60 0.00 2024.00 306.67 0.20 30.36 2.67 1.76
sdam 0.00 232.60 0.00 5.20 0.00 1902.40 365.85 0.17 31.77 2.81 1.46
sdby 0.00 391.60 0.00 8.40 0.00 3200.00 380.95 0.47 55.52 2.98 2.50
sddk 0.00 391.40 0.00 8.60 0.00 3200.00 372.09 0.44 51.23 2.72 2.34
(total=) 1.28
:
dm-18 0.00 0.00 0.00 1290.80 0.00 10326.40 8.00 58.64 45.43 0.02 2.58- sample #2: await time < service times
- From iostat man page:
"await
The average time (in milliseconds) for I/O requests issued to the device to be served. This
includes the time spent by the requests in queue and the time spent servicing them.svctm
The average service time (in milliseconds) for I/O requests that were issued to the device.
"
Device: rrqm/s wrqm/s r/s w/s rsec/s wsec/s avgrq-sz avgqu-sz await svctm %util
sdy 0.00 0.00 1.20 0.00 44.71 0.00 37.33 3.12 805.17 833.50 99.82
sdah 0.00 0.00 2.99 0.00 73.05 0.00 24.40 1.94 173.60 296.73 88.84
sdaw 0.00 0.20 1.20 0.40 35.13 4.79 25.00 4.65 20.00 568.00 90.70 Environment
- Red Hat Enterprise Linux 5.3z (2.6.18-128.1.14.el5)
Subscriber exclusive content
A Red Hat subscription provides unlimited access to our knowledgebase of over 48,000 articles and solutions.
Welcome! Check out the Getting Started with Red Hat page for quick tours and guides for common tasks.
