Getting "a start job is running for dev-disk-by..."

Ever had your Linux machine stall for a minute and a half on every boot, showing a message like this?

[  *** ] A start job is running for /dev/disk/by-uuid/abc12345-... (1min 30s / no limit)

It eventually carries on and boots fine — but that 90-second wait every single time gets old fast. This happened to me after I repartitioned my hard drive and recreated my swap.

What's actually going on

When you reformat or recreate a partition, Linux gives it a brand-new UUID (the unique ID that identifies the filesystem). But your /etc/fstab — the file that tells the system what to mount at boot — is still pointing at the old UUID. So at startup, systemd waits for a device that no longer exists, times out after 90 seconds, and moves on. The usual culprit is the swap entry.

Step 1 — Find the correct UUID

List your partitions and their current UUIDs:

sudo blkid

You'll get something like this — note the line whose TYPE is "swap":

/dev/sda1: UUID="9f3b...c2a1" TYPE="ext4" PARTUUID="..."
/dev/sda2: UUID="d41d8cd9-8f00-b204-e980-0998ecf8427e" TYPE="swap"

That swap UUID — d41d8cd9-...-0998ecf8427e in the example — is the correct, current one you want fstab to use.

Step 2 — Compare it with /etc/fstab

Open fstab:

sudo nano /etc/fstab

Find the swap line. The problem is a mismatch — the UUID here doesn't match what blkid just reported:

# before — points at the OLD swap UUID (the one causing the hang)
UUID=11111111-2222-3333-4444-555555555555 none swap sw 0 0

Replace that UUID with the current one from Step 1:

# after — now matches the real swap partition
UUID=d41d8cd9-8f00-b204-e980-0998ecf8427e none swap sw 0 0

Save and exit (in nano: Ctrl+O, Enter, Ctrl+X).

Note: Some systems reference swap by device path (e.g. /dev/sda2) instead of UUID. That works too, but UUIDs are safer — a device path like /dev/sda2 can change if you add or reorder drives, while the UUID follows the partition.

Step 3 — Test it without rebooting

You don't have to reboot to check your work. Reload the fstab definitions and turn swap on:

sudo systemctl daemon-reload

sudo swapon -a

If there's no error, confirm swap is active:

swapon --show

A populated table (showing your swap partition, its size, and how much is used) means fstab and the real partition now agree. You can also check with free -h — the Swap row should show your total.

Step 4 — Reboot

Reboot to confirm the hang is gone:

sudo reboot

The boot should now sail past that step instead of waiting 90 seconds.

Hope this helps!