2

Receiving the following information about an error encountered during system startup in dmesg. It wouldn't be a big deal, except it keeps triggering my log monitoring script! :)

[ 1.919823] PM: Error -22 checking image file
[ 1.919826] PM: Resume from disk failed.

Internet searching yields information suggesting it is related to 'hibernation/suspend' and 'apci' but this is server installation and 'hibernation/suspend' features were not included. (at least not deliberately included).

So, does something need to be blacklisted? The only real solution I've found on the web involves remaking the swap partition...which seems strange. Ideas?

Notes:
- I've installed Debian Squeeze with the minimal package set (expert install)
- I've seen this on multiple systems with different hardware.
- One a system with similar hardware but using the squeeze-backports kernel (2.6.39-bpo.2-amd64), does not produce this error
- I've seen this on i386 and amd64 targets

Aquadisco
  • 130
  • 8

1 Answers1

3

Every time kernel starts it tries to use swap partition to load its saved state (from suspend to disk). If there was no state saved, then it will produce such error. It's normal and harmless.

I'd suggesting to make your monitoring script to ignore this specific message (don't know if you're using a custom solution or a ready product, but you'll have much more messages to add to ignore list, so this will be a good exercise either way).

Hubert Kario
  • 6,361
  • 6
  • 36
  • 65
  • 1
    With that in mind, i then followed the [instructions](http://forums.techarena.in/operating-systems/1374672.htm) i alluded to earlier which recommend remaking the swap (mkswap) and the error disappeared. If it comes back, i can always filter it using Logwatch's ignore.conf as you suggested. – Aquadisco Jan 04 '12 at 14:10