[ovirt-users] snapshot create fails - files not owned by vdsm:kvm

I've got an issue with one of my oVirt 4.3.10.4 clusters. OS is Centos 7.7, storage is glusterfs. Whenever I try to create a snapshot of any of my VMs, I get an (eventual) error: VDSM track04.yard.psc.edu command HSMGetAllTasksStatusesVDS failed: Could not acquire resource. Probably resource factory threw an exception.: () After a bit of investigation, I believe that the root cause is that the snapshot file created is not owned by vdsm:kvm. Looking at the directories in glusterfs, I see that some of the disk images are owned by root:root, some are owned by qemu:qemu. In fact, if we watch the directory for the VM, we can actually see that the files are owned by vdsm:kvm when they are created, then get changed to qemu:qemu, then eventually get changed to being owned by root:root. This is entirely repeatable. Needless to say, oVirt can't read the disk images when they are owned by root, so that explains why the snapshot is failing. The question, then, is why the ownership is getting changed out from under the creation process. Checking the gluster volume info shows: Volume Name: engine Type: Replicate Volume ID: 00951055-74b5-463a-84c0-59fa03be7478 Status: Started Snapshot Count: 0 Number of Bricks: 1 x 3 = 3 Transport-type: tcp Bricks: Brick1: 10.200.0.131:/gluster_bricks/engine/engine Brick2: 10.200.0.134:/gluster_bricks/engine/engine Brick3: 10.200.0.135:/gluster_bricks/engine/engine Options Reconfigured: performance.client-io-threads: off nfs.disable: on transport.address-family: inet performance.quick-read: off performance.read-ahead: off performance.io-cache: off performance.low-prio-threads: 32 network.remote-dio: off cluster.eager-lock: enable cluster.quorum-type: auto cluster.server-quorum-type: server cluster.data-self-heal-algorithm: full cluster.locking-scheme: granular cluster.shd-max-threads: 8 cluster.shd-wait-qlength: 10000 features.shard: on user.cifs: off storage.owner-uid: 36 storage.owner-gid: 36 network.ping-timeout: 30 performance.strict-o-direct: on cluster.granular-entry-heal: enable I see that the owner UID and GID are correct (vdsm:kvm). Note that snapshots worked at one point in the past - I see that there are snapshot images from a year ago. Any ideas where to look to correct this? Thanks! Tod Pike