Error in operations-center provisioning cluster resync

IncusOS
1

I built a new operations-center last week and have created a couple clusters, one physical and one virtual, and a few days ago I noticed an error when resync’ing. The terraform-configuration artifacts for the clusters are not updating as of about the same time.

OC is keeping up-to-date with inventory changes in the two Incus clusters, however.

Here are the errors in the CLI:

$ ./operations-center/bin/operations-center.linux.amd64 provisioning cluster resync vc0
Received an error from the server: Failed to resync inventory for cluster: Failed to fetch instance "a097b0581d8b71398845337674879a665c3fc026a6fb5c5982e6e55d3be77830" in project "default": Instance not found
$ ./operations-center/bin/operations-center.linux.amd64 provisioning cluster resync pc0
Received an error from the server: Failed to resync inventory for cluster: Failed to fetch instance "7ffa05503866be99a7fc2ebe865bc061767af1d4922dc5a3d62f83b43d81a95f" in project "default": Instance not found

I have no idea what those hex numbers represent.

Here are the stuck terraform artifacts which would identify the approximate time of the first resync error:

$ ./operations-center/bin/operations-center.linux.amd64 provisioning cluster artifact list vc0 --format compact
          Name            Cluster                         Description                                 Last Updated
 terraform-configuration  vc0      Initial terraform configuration used for post-clustering.  2025-12-14 19:05:51 +0000 UTC
$ ./operations-center/bin/operations-center.linux.amd64 provisioning cluster artifact list pc0 --format compact
          Name            Cluster                         Description                                 Last Updated
 terraform-configuration  pc0      Initial terraform configuration used for post-clustering.  2025-12-14 19:25:47 +0000 UTC

The oc client version was 0.0.1 (git clone GitHub - FuturFusion/operations-center: FuturFusion Cloud Operations Center main branch as of November 18) but I first saw the error in the webui of OC which was fairly current as of the error. Using today’s main of the oc client gives the same error (version oddity; it looks like main is based off tags/0.1.0 but the built executable still says 0.0.1 for the version).

OC itself appears current (looking at its virtual console, but I notice there is no easy way to see the version in the webui). The OC application console version string is 202512131701, as is the underlying IncusOS.

2

That issue has been fixed in Incus already. A fix is in the current testing images for IncusOS I believe, but we’re looking at publishing a new stable image later today.

1 Like
3

Sweet, I’ll test when it comes down! Thanks! I’ll also try to look in the main Incus repo for matching issues in the future.

4

I just tried to search for ‘sync’ and ‘resync’ in the GitHub - lxc/incus: Powerful system container and virtual machine manager repo under issues and PRs and also skim-searched commit messages from the last week. What search would have found the fix given my naive ones didn’t seem to?

5

It wouldn’t have been easy to find it from the Operations Center error :slight_smile:

The fix is: incusd/storage_volumes: Fix state handling in getVolumeFull · lxc/incus@4bda52a · GitHub

6

Ah, cool, I don’t have feel bad about my searching skills. :wink:

1 Like