Recently, I performed a dist-upgrade from Debian Wheezy to Jessie. The upgrade was performed without any error but after the upgrade the swap partition was no longer mounted/activated. The mentioned system is a KVM (qemu-kvm, centos7).
During the start up of the system, systemd tries to mount the swap partition but fails due to timeout with the message “Dependency failed for…”
What I tried so far:
- performing “swapon /dev/sda2” manually works without any problem
- recreated the swap partition with “mkswap /dev/sda2”
- double check fstab entry and used UUID
- removed the fstab entry so that systemd detects it automatically
- different option parameter (fstab)
- asked the german debian community -but no helpful response: https://debianforum.de/forum/viewtopic.php?f=32&t=159902&p=1084179#p1084179
- manually mounting with systemd (systemctl start dev-disk…) fails with the same error but after perfoming “swapon” and “swapoff” it works.
What I found so far:
- actually, there is a closed ticket which describes my problem very
well : https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=764539
-however, the problem was due to a custom kernel (missing options) -and I do not use any custom kernel so the mentioned options are enabled
- there is a serverfault thread where a user describes a
similiar problem but had an other system and setup -and without any
Swap partition not mounted on boot
journalctl -r | grep “dev-disk”
Mar 07 22:31:37 hostname systemd: Dependency failed for /dev/disk/by-uuid/062c87ce-135e-45fc-b430-7336a9ab164a. Mar 07 22:31:37 hostname systemd: Timed out waiting for device dev-disk-by\x2duuid-062c87ce\x2d135e\x2d45fc\x2db430\x2d7336a9ab164a.device. Mar 07 22:31:37 hostname systemd: Job dev-disk-by\x2duuid-062c87ce\x2d135e\x2d45fc\x2db430\x2d7336a9ab164a.device/start timed out.
UUID=5443f0ed-f437-4804-8fa0-45a039413b9a / ext4 errors=remount-ro 0 1 UUID=062c87ce-135e-45fc-b430-7336a9ab164a none swap sw 0 0
Linux hostname 3.16.0-4-amd64 #1 SMP Debian 3.16.7-ckt20-1+deb8u4 (2016-02-29) x86_64 GNU/Linux
/dev/sda1: UUID="5443f0ed-f437-4804-8fa0-45a039413b9a" TYPE="ext4" PARTUUID="0006f6b9-01" /dev/sda2: UUID="062c87ce-135e-45fc-b430-7336a9ab164a" TYPE="swap" PARTUUID="0006f6b9-02"
parted /dev/sda – print
(parted) print Model: ATA QEMU HARDDISK (scsi) Disk /dev/sda: 10.7GB Sector size (logical/physical): 512B/512B Partition Table: msdos Disk Flags: Number Start End Size Type File system Flags 1 1049kB 10.2GB 10.2GB primary ext4 boot 2 10.2GB 10.7GB 491MB primary linux-swap(v1)
dpkg -l systemd
ii systemd 215-17+deb8u3 amd64 system and service manager
ls -l /dev/disk/by-uuid
lrwxrwxrwx 1 root root 10 Mar 24 23:01 062c87ce-135e-45fc-b430-7336a9ab164a -> ../../sda2 lrwxrwxrwx 1 root root 10 Mar 20 18:44 5443f0ed-f437-4804-8fa0-45a039413b9a -> ../../sda1
The problem is in your
UUID=062c87ce-135e-45fc-b430-7336a9ab164a none swap sw 0 0
This format refers to a block device by UUID, which is stored in the metadata of the swap space (or other filesystem). Because you recreated the swap space with
mkswap, the UUID is no longer valid. More to the point, because upgrading Debian caused the swap to be recreated, the UUID is no longer valid.
You can resolve this in one of two ways:
- Update the UUID in
/etc/fstabto the current UUID of the swap device. You can find this with the
- Refer to the device by its
/devnode rather than UUID, i.e.
/dev/sda2. This is not recommended, as the device node may change if you add or remove storage devices.
This work is licensed under a Creative Commons Attribution-ShareAlike 3.0 Unported License.