# ZFS startup fails after Rocky Linux 8 reinstallation -- need a hint

**URL:** <https://forum.45homelab.com/t/zfs-startup-fails-after-rocky-linux-8-reinstallation-need-a-hint/3334>\
**Category:** General\
**Tags:** houston, myhomelab, help-and-suggestions\
**Created:** [May 29, 2025, 2:55pm UTC](https://forum.45homelab.com/t/zfs-startup-fails-after-rocky-linux-8-reinstallation-need-a-hint/3334 "2025-05-29T14:55:28Z")\
**Posts on this page:** 3\
**Page:** 1

<div class="post-metadata">

**Author:** ![DismalWizard](https://forum.45homelab.com/letter_avatar_proxy/v4/letter/d/94ad74/32.png) [@DismalWizard](https://forum.45homelab.com/u/DismalWizard)\
**Post date:** [May 29, 2025, 2:55pm UTC](https://forum.45homelab.com/t/zfs-startup-fails-after-rocky-linux-8-reinstallation-need-a-hint/3334/1 "2025-05-29T14:55:28Z")

</div>

Somebody rushed a system update interrupting it before completion. The result was not bootable. So I reinstalled Rocky 8 and ran the 45drives scripts. End result is that ZFS is there but won’t mount the pool at boot. The complaint that zpool command is missing is incorrect.

At first, I thought there was a dinged configuration item. Inspecting the zfs-import-scan.service file, it looked unremarkable. Clicking the restart button for the service, it starts and ZFS pool mounts as expected.

At this point it still needs zfs mount -a issued.

Once mounts are complete, containers can be started.

At second though, is there a race? Or a wait-for-something returning early?  
DVR storage is not there until the pool starts. (Channels from FancyBits).

Or at third thought, a premature release from a step preceding ZFS startup? I need a hint for what should be checked next.

Here’s the log entry for zfs-import-scan service, timestamps deleted  
/usr/local/sbin/zpool is there, alive and well. Assume it is doing something exotic like zpool import -a which works. Without the pool there, my DVR doesn’t come up and other annoyances.

Failed to start Import ZFS pools by device scanning.

zfs-import-scan.service: Failed with result ‘exit-code’.

zfs-import-scan.service: Main process exited, code=exited, status=203/EXEC

zfs-import-scan.service: Failed at step EXEC spawning /usr/local/sbin/zpool: No such file or directory

zfs-import-scan.service: Failed to execute command: No such file or directory

---

<div class="post-metadata">

**Author:** ![DismalWizard](https://forum.45homelab.com/letter_avatar_proxy/v4/letter/d/94ad74/32.png) [@DismalWizard](https://forum.45homelab.com/u/DismalWizard)\
**Post date:** [May 29, 2025, 7:27pm UTC](https://forum.45homelab.com/t/zfs-startup-fails-after-rocky-linux-8-reinstallation-need-a-hint/3334/2 "2025-05-29T19:27:58Z")

</div>

I’m impressed by Cockpit! Wow. It lets you crawl the services graph!  
The logs are linked to services entries  
The systemd startup files are declaritive and pretty easy to read.

I fiddled with udev.conf to enable info level logging and rebooted.  
Nothing untoward appeared in the logs. Timing changes from disk-filling logging didn’t help

Welcome to Wilmington, NC. Those of you who are good can work in WIlmington in the winter! 🙂 Keep ears open for tick-tock… tick-tock… tick-toc … (Captain Hook’s buddy). There are alligators as far north as Wilmington. 😵‍💫

---

<div class="post-metadata">

**Author:** ![DismalWizard](https://forum.45homelab.com/letter_avatar_proxy/v4/letter/d/94ad74/32.png) [@DismalWizard](https://forum.45homelab.com/u/DismalWizard)\
**Post date:** [June 25, 2025, 1:39pm UTC](https://forum.45homelab.com/t/zfs-startup-fails-after-rocky-linux-8-reinstallation-need-a-hint/3334/3 "2025-06-25T13:39:46Z")

</div>

I found the scan for pools service off. Changed to run and life is good.