A Start Job Is Running For dev-sda1

[DougTheGuitarGuy]

I had Linux Mint 18 on this box and created a RAID 0 stripe with mdadm using two 2TB drives connected to a Promise Fasttrak TX4310 controller which I let Linux define the array and the Promise card just just a dummy sata interface, no array defined.

On installing GeckoLinux, I unplugged the card to avoid any GRUB errors, as GRUB likes to stick boot files in the wrong places sometimes.  After installing mdadm in GeckoLinux, and running the mdadm --scan command I noticed the array had already been recognized after rebooting, and gecko will mount the array and access the files, it still shows it was created in Linux Mint 18 in the Partitioner applet, and identifies it as /dev/md/P5KPL-LinuxMint18:0

I was tickled to see the array had migrated safely to the new install!

Question now, on boot I get a 1min 30 second timeout while the system says "a start job is running for dev-sda1.device" where the two drives in the array are sda and sdb.  What would be the purpose of this start job and is there a way to fix the 90 second timeout on boot?

[S. B.]

Hi there, this thread appears to be a similar issue:
https://forums.opensuse.org/showthread.php/503587-Slow-boot-What-is-quot-A-start-job-is-running-for-dev-disk-by-quot

What is /dev/sda1 used for? Swap partition for your GeckoLinux install possibly?

Also you can run this command to see exactly what is holding up the boot process:
sudo systemd-analyze critical-chain

[DougTheGuitarGuy]

I got a screen cap with my phone (lol) right after the wait period, and it said:

[TIME] Timed out waiting for device dev-sda1.device.
[DEPEND] Dependency failed for Resume from hibernation using device /dev/sda1.

This is a RAID 0 stripe that has a single partition or volume on the disc that says linux-raid for /dev/sda and the same thing for the other volume /dev/sdb  there is no /dev/sda1 on the system.

I did not have the RAID connected when I ran the install for gecko linux.  Now that it is connected, device drive letters have changed.

My boot drive is an SSD that has now become /dev/sdc and it has 3 partitions on it, sdc1 being the swap partition, so at the time of Gecko Linux install, /dev/sdc WAS /dev/sda and this is what the timer is waiting for so how can I edit the file that has the path for hibernation?  If I can change /dev/sda1 to /dev/sdc1 it would then think it is resuming from hibernation correctly and the 90sec timer would go away, right?

 

[S.]

OK, that makes sense. Could you please post your /etc/fstab ? I think
it's in there.

[DougTheGuitarGuy]

UUID=de68daf8-22a1-4eb6-a31b-f0f42d7a665a swap swap defaults 0 0
UUID=48e97fa7-5b56-4347-87ec-0d99e8430ee4 / btrfs defaults 0 0
UUID=48e97fa7-5b56-4347-87ec-0d99e8430ee4 /boot/grub2/i386-pc btrfs subvol=@/boot/grub2/i386-pc 0 0
UUID=48e97fa7-5b56-4347-87ec-0d99e8430ee4 /boot/grub2/x86_64-efi btrfs subvol=@/boot/grub2/x86_64-efi 0 0
UUID=48e97fa7-5b56-4347-87ec-0d99e8430ee4 /opt btrfs subvol=@/opt 0 0
UUID=48e97fa7-5b56-4347-87ec-0d99e8430ee4 /srv btrfs subvol=@/srv 0 0
UUID=48e97fa7-5b56-4347-87ec-0d99e8430ee4 /tmp btrfs subvol=@/tmp 0 0
UUID=48e97fa7-5b56-4347-87ec-0d99e8430ee4 /usr/local btrfs subvol=@/usr/local 0 0
UUID=48e97fa7-5b56-4347-87ec-0d99e8430ee4 /var/cache btrfs subvol=@/var/cache 0 0
UUID=48e97fa7-5b56-4347-87ec-0d99e8430ee4 /var/crash btrfs subvol=@/var/crash 0 0
UUID=48e97fa7-5b56-4347-87ec-0d99e8430ee4 /var/lib/libvirt/images btrfs subvol=@/var/lib/libvirt/images 0 0
UUID=48e97fa7-5b56-4347-87ec-0d99e8430ee4 /var/lib/machines btrfs subvol=@/var/lib/machines 0 0
UUID=48e97fa7-5b56-4347-87ec-0d99e8430ee4 /var/lib/mailman btrfs subvol=@/var/lib/mailman 0 0
UUID=48e97fa7-5b56-4347-87ec-0d99e8430ee4 /var/lib/mariadb btrfs subvol=@/var/lib/mariadb 0 0
UUID=48e97fa7-5b56-4347-87ec-0d99e8430ee4 /var/lib/mysql btrfs subvol=@/var/lib/mysql 0 0
UUID=48e97fa7-5b56-4347-87ec-0d99e8430ee4 /var/lib/named btrfs subvol=@/var/lib/named 0 0
UUID=48e97fa7-5b56-4347-87ec-0d99e8430ee4 /var/lib/pgsql btrfs subvol=@/var/lib/pgsql 0 0
UUID=48e97fa7-5b56-4347-87ec-0d99e8430ee4 /var/log btrfs subvol=@/var/log 0 0
UUID=48e97fa7-5b56-4347-87ec-0d99e8430ee4 /var/opt btrfs subvol=@/var/opt 0 0
UUID=48e97fa7-5b56-4347-87ec-0d99e8430ee4 /var/spool btrfs subvol=@/var/spool 0 0
UUID=48e97fa7-5b56-4347-87ec-0d99e8430ee4 /var/tmp btrfs subvol=@/var/tmp 0 0
UUID=48e97fa7-5b56-4347-87ec-0d99e8430ee4 /.snapshots btrfs subvol=@/.snapshots 0 0
UUID=e5a3383b-6b68-4199-99cc-6409061f18ba /home                xfs        defaults              1 2

[S.]

Thanks! If I'm not mistaken the resume device can be changed via the
YaST Boot Loader module. Could you look in there under the "Kernel
Parameters" tab, and try to relate the device listed there to the fstab
you posted? It might have the UUID of the device listed instead of the
/dev/sdX name.

[DougTheGuitarGuy]

Yup, here is what is/was in there:

resume=/dev/sda1 splash=silent quiet showopts

[DougTheGuitarGuy]

And here is what is in there now:

resume=/dev/sdc1 splash=silent quiet showopts

Let me reboot and test it.  Apparently the boot loader saved the change.

[S.]

Great, so just change the resume= parameter to the proper resume device
with your new setup. (Sounds like /dev/sdc1 from what you've said
previously?)

[DougTheGuitarGuy]

Yes and the issue is FIXED!

[S.]

Sweet, glad you were able to figure it out!

[DougTheGuitarGuy]

I updated the thread you posted earlier as well.

Thanks!