sunbeam bootstrap microceph Error: {'return-code': 0}

Bug #2031157 reported by Johannes Jungkunst
22
This bug affects 5 people
Affects Status Importance Assigned to Milestone
OpenStack Snap
New
Undecided
Unassigned

Bug Description

during sunbeam bootstrapping, the installation always stops at the microceph step:

ubuntu@sunbeam01:~$ sunbeam cluster bootstrap --role control --role compute --role storage
...
Disks to attach to MicroCeph (/dev/disk/by-id/nvme-eui.e8238fa6bf530001001b448b4e330522):
Error: {'return-code': 0}

Revision history for this message
Matt Verran (mv-2112) wrote :

If you are 100% sure that's the disk you want to use for Ceph try wipefs -af /dev/disk/by-id/nvme-eui.e8238fa6bf530001001b448b4e330522 [THIS WILL DESTROY YOUR DATA ON THAT DEVICE]

Its also worth noting if you use multiple devices its comma separated although its not mentioned anywhere (might be nice to be space or comma?).

Revision history for this message
Chi Wai CHAN (raychan96) wrote :

I've also encountere the same issue here. In my case, the disk that I used was previously used by microceph, and it didn't get wiped out properly. The error I am seeing is shown in the other bug report: https://bugs.launchpad.net/charm-microceph/+bug/2038504.

Revision history for this message
Scott Storck (scttstrck) wrote :

When my lab stops with the error in the bug description, then it usually due to that microceph is already installed from a previous attempt and is currently using the disks, or the disks weren't cleaned up.

You can check if microceph is currently using the disks by running:
sudo microceph disk list

The output will contain two blocks.

Disks configured in MicroCeph:
+-----+----------+------+
| OSD | LOCATION | PATH |
+-----+----------+------+

Available unpartitioned disks on this system:
+-------------+-----------+------+----------------------------------------+
| MODEL | CAPACITY | TYPE | PATH |
+-------------+-----------+------+----------------------------------------+

If the device is listed in the first block, then you need to remove it from microceph with the following command:
sudo microceph disk remove <OSDNUMBER> --bypass-safety-checks

If it is in the second block, or the command doesn't work, you need to clear it with wipefs:
sudo wipefs -a -f <DEVICEPATH>

You should see "bytes were erased at offset" in the output if it was successfully cleared.

To post a comment you must log in.
This report contains Public information  
Everyone can see this information.

Other bug subscribers

Remote bug watches

Bug watches keep track of this bug in other bug trackers.