'Clearing orphaned inodes' and state lost on hibernation resume

25,728

Well, I am surprised nobody suggested this as I have been having this problem for quite some time. The answer it seemed was staring right in my face. Apparently my swap partition was roughly the same size as my memory. Also, I hadn't added the link to my swap partition's UUID in my grub file. After increasing the swap partition size to double that of the memory and adding its UUID in the grub file the hibernation resume is working normally for the past few days. Although, the resume from hibernation takes a tad bit longer but I'm not complaining.

You need to make sure that your swap partition is defined in the following files:

  • /etc/initramfs-tools/conf.d/resume
  • /etc/default/grub

UPDATE

Using low level interface uswsusp as default hibernating mechanism drastically improved my resume time to under a minute !!!

sudo apt-get install uswsusp

Create file /etc/pm/config.d/00sleep_module and add following line:

  • SLEEP_MODULE="uswsusp"
Share:
25,728

Related videos on Youtube

thethakuri
Author by

thethakuri

Updated on September 18, 2022

Comments

  • thethakuri
    thethakuri over 1 year

    I installed a fresh 16.04 LTS. I have been experiencing some problems with Wifi applet display error (after resume from Suspend) and with hibernation resume. I enabled hibernation on menu using the method shown here . Now hibernation resume is not working intermittently. Sometimes it works fine, other times it displays text on resume that says something about 'clearing orphaned inodes' and the system simply boots up anew, without the prior memory state.

    Here's some info:

    $ sudo blkid
    /dev/sda1: LABEL="System Reserved" UUID="50921EE4921ECE7A" TYPE="ntfs" PARTUUID="dda192f8-01"
    /dev/sda2: LABEL="Primary Disk" UUID="765E305F5E3019F7" TYPE="ntfs" PARTUUID="dda192f8-02"
    /dev/sda3: LABEL="Secondary Disk" UUID="E2D42C6AD42C42E1" TYPE="ntfs" PARTUUID="dda192f8-03"
    /dev/sda5: UUID="dbaad068-46da-4637-9c45-5c32c20d3cfe" TYPE="swsuspend" PARTUUID="dda192f8-05"
    /dev/sda6: UUID="31385b29-f351-4a10-9dcf-c92efd58334b" TYPE="swap" PARTUUID="dda192f8-06"
    /dev/sda7: UUID="1f734f56-7328-4029-88a0-fa995426d4d2" TYPE="ext4" PARTUUID="dda192f8-07"
    
    $ cat /etc/initramfs-tools/conf.d/resume
    RESUME=UUID=31385b29-f351-4a10-9dcf-c92efd58334b
    
  • jawtheshark
    jawtheshark over 7 years
    I'm not saying this is not the solution, but the odds of having a broken/empty CMOS battery are rather low. I would warn against just assuming such exotic problems.
  • Victor - Reinstate Monica
    Victor - Reinstate Monica over 7 years
    That is basically why I wrote detailed steps for debugging, not just my problem. It could well be a non-reallocatable bad sector in the swap partition, the mentioned dualboot timezone nosystemd problem - or even the "evil maid attack" hypothesized above. (Not many other things to think about, though - if it works sometimes.)
  • Victor - Reinstate Monica
    Victor - Reinstate Monica over 7 years
    The other thing to watch for, if you do not use systemd, that your resume (swap) device may switch over so that kernel will not find the resume image. If this is the problem, you may use blkid to find the UUID of your swap partition then add resume=UUID=<the uuid here> to the kernel config file in grub.cfg, just as the debug params added above.
  • thethakuri
    thethakuri over 7 years
    What'd you know. My CMOS battery was dying as the time would reset on each boot of W7 (Ubuntu time wasn't affected though). I have replaced the old battery with a new one. Give me few days to test it out.
  • Victor - Reinstate Monica
    Victor - Reinstate Monica about 7 years
    Is it still working? My swap was 1.5x the size of my physical memory, and it did that sometimes, and did not at others.