- Users - oVirt List Archives

Re: Hosted Engine Deploy Failure
by Mihai Puha 31 Jan '24

31 Jan '24

Hello Gary, How did you found that there were 2 engine VMs competing? Mihai,

1 0

Cannot remove Snapshot. The VM is during a backup operation.
by and＠missme.ro 30 Jan '24

30 Jan '24

Hello! Running ovirt Version 4.5.5-1.el8 I had an issue with the iscsi server during the backup and I have two VMs that cannot be backed up anymore by Veeam. In the ovirt event log i have the following errors: Snapshot 'Auto-generated for Backup VM' creation for VM 'dns-a' has been completed. VDSM ovirt1-02 command StartNbdServerVDS failed: Bitmap does not exist: "{'reason': 'Bitmap does not exist in /rhev/data-center/mnt/blockSD/b2fa3469-a380-4180-a89a-43d65085d1b9/images/6a4de98a-b544-4df8-beb1-e560fd61c0e6/cdb26b8b-c447-48de-affa-d7f778aebac7', 'bitmap': '12d2fb20-74da-4e63-b240-f1a42210760c'}" Transfer was stopped by system. Reason: failed to create a signed image ticket. Image Download with disk dns-a_Disk1 was initiated by veeam@internal-authz Image Download with disk dns-a_Disk1 was cancelled. The error on the Veeam backup proxy: dns-a: Unable to create image transfer: Reason: 'Operation Failed', Detail: '[]' When trying to delete the snapshot from the administration interface I receive the following error in the web interface (and nothing gets logged in the event log) Cannot remove Snapshot. The VM is during a backup operation. How should I go about fixing this issue?

2 1

HE Storage Domain Path Config Setting - Where?
by Matthew J Black 30 Jan '24

30 Jan '24

Hey Guys, Quick Q: In which file (on a Hosted-Engine or Hosted-Engine Host) is the configuration for the path to a Storage Domain kept - in particular, the "hosted-engine" Storage Domain? I've got something "funny" going on: the logs (as far as I can see) are reporting that 2 of my 3 HE-Hosts can't connect to the HE Storage Domain (but don't explain why), and the OVE GUI is reporting an "odd" (ie incorrect; non-existent) path to the HE Storage Domain. Via CLI I have confirmed that all three HE Hosts *can* reach (ie have the correct "findmnt" mappings) to the HE Storage Domain's actual file location, and I can't locate the "ghost" HE Storage Domain path or its config setting anywhere - so I don't even know if that's the issue, but I'd like to eliminate it from my trouble-shooting process. Anyway, if someone could get back to me, please, I'd really appreciate it. Cheers Dulux-Oz

2 1

Geo-replication configuration problem
by Ismet Sonmez 28 Jan '24

28 Jan '24

Hello, Newly installed node version 4.5.5 2 clusters and three nodes each every time I try to replicate the geo in cluster 1 it gives an error 3 errors like this: VDSM node4 command UpdateGlusterGeoRepKeysVDS failed: Internal JSON-RPC error: {'cause': 'Attempting to call function: GlusterVolume.geoRepKeysUpdate bound method of object <vdsm.gluster.apiwrapper.GlusterVolume at address 0x7f1524534f28>> with arguments: (\'root\', \'command="/usr/libexec/glusterfs/gsyncd" ssh-rsa AAAAB3N**** \'command="tar ${SSH_ORIGINAL_COMMAND#* }" ssh-rsa AAAA**** \'command="/usr/libexec/glusterfs/gsyncd" ssh-rsa AAAAB3**** \'command="tar ${SSH_ORIGINAL_COMMAND#* }" ssh-rsa AAAA*** \'command="/usr/libexec/glusterfs/gsyncd" ssh-rsa AAAA******* \'command="tar ${SSH_ORIGINAL_COMMAND#* }" ssh-rsa AAAA**** ]) error: bytes-like object required, not \'str\' } I understand that when creating the rsa key, or sending it to the node. How can I solve it?

1 0

Forbid hosts/nodes from assembling a soft-raid that was created inside a VM
by Vladislav Solovei 27 Jan '24

27 Jan '24

When I create a soft-raid (md-raid) inside a virtual machine (yes, sometimes it's necessary to do this to get a disk inside the VM with a capacity of more than 8TB), the host/node detects this array and assembles it automatically. I tried adding the parameter 'raid=noautodetect' to the kernel boot parameters, but it doesn't help. Is it possible to prevent hosts from doing this? :)

1 0

Low confirmed free space on gluster volume
by Jonas 27 Jan '24

27 Jan '24

Hello list I am regularly getting the following error on a Gluster volume hosted on a three node hyperconverged oVirt-Cluster: Warning! Low confirmed free space on gluster volume tier1-owncloud-users-01 This volume is configured with 3TiB (2.5TIB used) and is used as the name implies to store data of an ownCloud instance. It is not used as a storage domain in oVirt. I don't really want to resize the volume just to make the warning go away. While clicking through the oVirt web interface i found the configuration "Warning Low Confirmed Space Indicator" in the advanced parameters of the storage domains but did not find any similar setting in the configuration of the gluster volumes. Do you know of way to configure this setting for a gluster volume which is not used as a storage domain? Thank you Jonas

1 0

Change Hosted Engine VM MAC address
by Sergei Panchenko 26 Jan '24

26 Jan '24

Good morning, colleagues! Due some network issues I need to change HostedEngine VM MAC address. The additional diffeculty is unaccessebility of the HE Web-interface via the network (the cause of - those network issues -)). Is there any way to change the HE vNIC's MAC using command line (on the host where HE VM works or on the HE VM)? Thanks in advance, Sergei.

2 2

Need help again, new issue with VM import
by Michaal R 24 Jan '24

24 Jan '24

I have a VM that's a little over 4TB in size that won't import for whatever reason. I've tried changing the export from the streaming disk format to a flat vmdk,thinking that might be it, but it didn't work. I've gone through the OVF and don't see anything that stands out. Same for the logs, I see where it's having issues importing the 4TB drive from the VM, but I can't decipher the error messages well enough to know how to fix it. The drives had chkdsk run on them before they were exported, and the VM was cleanly shut down, so I don't understand what the issue is. Not even sure at this point how to run a filesystem repair on the vmdk, in case the export corrupted something. And I can't export it directly from ESXi either, for some reason. Each attempt fails with an unspecified error. Below is a snip from the import log where it's reading the drives. It has one of the first errors: [ 2.492229] scsi host0: Virtio SCSI HBA [ 2.501365] scsi 0:0:0:0: Direct-Access QEMU QEMU HARDDISK 2.5+ PQ: 0 ANSI: 5 [ 2.504080] scsi 0:0:1:0: Direct-Access QEMU QEMU HARDDISK 2.5+ PQ: 0 ANSI: 5 [ 2.506035] scsi 0:0:2:0: Direct-Access QEMU QEMU HARDDISK 2.5+ PQ: 0 ANSI: 5 [ 2.540737] sd 0:0:0:0: Power-on or device reset occurred [ 2.540790] sd 0:0:1:0: Power-on or device reset occurred [ 2.540899] sd 0:0:2:0: Power-on or device reset occurred [ 2.541130] sd 0:0:0:0: [sda] 8589934592 512-byte logical blocks: (4.40 TB/4.00 TiB) [ 2.541168] sd 0:0:2:0: [sdc] 8388608 512-byte logical blocks: (4.29 GB/4.00 GiB) [ 2.541319] sd 0:0:2:0: [sdc] Write Protect is off [ 2.541391] sd 0:0:0:0: [sda] Write Protect is off [ 2.541592] sd 0:0:2:0: [sdc] Write cache: enabled, read cache: enabled, doesn't support DPO or FUA [ 2.542019] sd 0:0:0:0: [sda] Write cache: enabled, read cache: enabled, doesn't support DPO or FUA [ 2.544060] sd 0:0:2:0: [sdc] Attached SCSI disk [ 2.545294] sd 0:0:1:0: [sdb] 251658240 512-byte logical blocks: (129 GB/120 GiB) [ 2.545437] sd 0:0:1:0: [sdb] Write Protect is off [ 2.545966] sd 0:0:1:0: [sdb] Write cache: enabled, read cache: enabled, doesn't support DPO or FUA [ 2.551250] sdb: sdb1 sdb2 [ 2.556698] sd 0:0:1:0qemu-nbd: Disconnect client, due to: Failed to send reply: reading from file failed: Invalid argument : [sdb] Attached SCSI disk qemu-nbd: Disconnect client, due to: Failed to send reply: reading from file failed: Invalid argument qemu-nbd: Disconnect client, due to: Failed to send reply: reading from file failed: Invalid argument [ 2.618883] sd 0:0:0:0: [sda] tag#157 FAILED Result: hostbyte=DID_OK driverbyte=DRIVER_OK cmd_age=0s [ 2.618890] sd 0:0:0:0: [sda] tag#157 Sense Key : Aborted Command [current] [ 2.618892] sd 0:0:0:0: [sda] tag#157 Add. Sense: I/O process terminated [ 2.618894] sd 0:0:0:0: [sda] tag#157 CDB: Read(16) 88 00 00 00 00 01 ff ff ff f8 00 00 00 08 00 00 [ 2.618897] I/O error, dev sda, sector 8589934584 op 0x0:(READ) flags 0x0 phys_seg 1 prio class 2 supermin: intern[ 2.618902] Buffer I/O error on dev sda, logical block 1073741823, async page read al insmod virtio[ 2.618927] Alternate GPT is invalid, using primary GPT. The rest of the log is peppered with Buffer I/O error entries on /dev/sda (the 4TB vmdk). Could someone please help? I think I've been looking at these logs and trying to fix the drive for so long I've gone logic blind and can't see the answer right in front of my eyes. Here's a link to the logs pulled from the host: https://www.dropbox.com/scl/fi/570t279s0k3pfgafuvv01/felicity-import-1.18.2…

1 0

Re: [External] : Re: can hosted engine deploy use local repository mirrors instead of internet ones?
by iucounu＠gmail.com 24 Jan '24

24 Jan '24

Hi Marcos, > > The dnsmasq service running on the KVM host manages the IP assignment during the first > deployment phase. > How did you deploy your KVM host? Which configurations have you done on it before running > the hosted-engine --deploy? > Also, what is your full hosted-engine deployment command? > I deployed the KVM hosts using the guide at: https://www.ovirt.org/documentation/installing_ovirt_as_a_self-hosted_engin… (section 4.2 Installing Enterprise Linux Hosts) and https://www.ovirt.org/download/install_on_rhel.html Most of the ovirt packages were installed from the installation of the ovirt-engine-appliance. The KVM hosts are using EL 9.3. I've tried on two different EL hosts: one with the standard ovirt-45 repos, and the other had the nightly builds enabled. I observed the same issues on both. I haven't changed any specific settings, such as networking or storage. The full command to deploy the engine VM was done via: hosted-engine --deploy --4 I have just setup glusterfs just as a temporary storage option now (yet to run the deploy again), though I'm not sure how to get ovirt to use it. As mentioned I don't know whether this or the networking is causing it to fail. In case it is important, I notice that virtnetworkd.socket systemd unit gets killed during the deployment, and has to be restarted otherwise the deployment fails prematurely. It also masks all the libvirtd systemd units as part of the cleanup, and these have to manually unmasked and several need manually restarting (virtnetword.socket, virtqemud.socket and virtstoraged.sock) before the deployment is run again or the deploy will fail on trying to communicate with these. Thanks very much for the help, let me know if any further information needed. Cam > > -----Original Message----- > From: iucounu(a)gmail.com <iucounu(a)gmail.com> > Sent: Tuesday, January 23, 2024 12:49 PM > To: users(a)ovirt.org > Subject: [ovirt-users] Re: [External] : Re: can hosted engine deploy use local repository > mirrors instead of internet ones? > > Thanks very for the reply Marcos. I tried another deployment, just to see if "wait > for the host to be up" would time out, and I saw a couple of errors in the log: > > From the ovirt-hosted-engine-setup-ansible-final_clean log, it mentions that the VM IP is > undefined: > > 2024-01-23 12:50:19,554+0000 DEBUG otopi.ovirt_hosted_engine_setup.ansible_utils > ansible_utils._process_output:109 {'msg': "The task includes an option with > an undefined variable. The error was: 'local_vm_ip' is undefined. > 'local_vm_ip' is undefined\n\nThe error appears to be in > '/usr/share/ansible/collections/ansible_collections/ovirt/ovirt/roles/hosted_engine_setup/tasks/sync_on_engine_machine.yml': > line 2, column 3, but may\nbe elsewhere in the file depending on the exact syntax > problem.\n\nThe offending line appears to be:\n\n---\n- name: Set the name for add_host\n > ^ here\n", '_ansible_no_log': False} > > In the ovirt-hosted-engine-setuplog, it mentions not being able to get the storage pool: > > > 2024-01-23 12:50:35,787+0000 DEBUG otopi.ovirt_hosted_engine_setup.ansible_utils > ansible_utils._process_output:109 {'changed': True, 'stdout': '', > 'stderr': "error: failed to get pool 'localvmvy8whst5'\nerror: > Storage pool not found: no storage pool with matching name > 'localvmvy8whst5'", 'rc': 1, 'cmd': ['virsh', > '-c', 'qemu:///system?authfile=/etc/ovirt-hosted-engine/virsh_auth.conf', > 'pool-destroy', 'localvmvy8whst5'], 'start': '2024-01-23 > 12:50:35.558666', 'end': '2024-01-23 12:50:35.611808', > 'delta': '0:00:00.053142', 'msg': 'non-zero return code', > 'invocation': {'module_args': {'_raw_params': 'virsh -c > qemu:///system?authfile=/etc/ovirt-hosted-engine/virsh_auth.conf pool-destroy > localvmvy8whst5', '_uses_shell': False, 'stdin_add_newline': True, > 'strip_empty_ends': True, 'argv': None, 'chdir': None, > 'executable': None, 'creates': None, 'removes': None, > 'stdin': None}}, 'stdout_lines': [], 'stderr_lines': ["error: > failed to get pool 'localvmvy8whst5'", "error: Storage pool not fou > nd: no storage pool with matching name 'localvmvy8whst5'"], > '_ansible_no_log': None} > > I set the IP to the one I have assigned in DNS, but when I attach to the console of the VM > (which is still running, though the disk image has been deleted) via virsh, it shows me a > completely different IP: In hosted-engine --deploy, I set a 10.0.0.x address, however, it > shows a 192.168.1.x address on the VM. Do I need to set this somewhere else, e.g., with > '--ansible-extra-vars=he_ipv4_subnet_prefix='? > > As for the storage pool, is that for later VM deployment? The deploy script did not ask me > for a storage location. If I need to specify this, where do I do this? > > Thanks again for any help, > > Kind regards, > > Cam > > PS: is there a simple way to have the answers saved so I don't have to keep running > through all the questions every time I try a deployment > _______________________________________________ > Users mailing list -- users(a)ovirt.org > To unsubscribe send an email to users-leave(a)ovirt.org Privacy Statement: > https://urldefense.com/v3/__https://www.ovirt.org/privacy-policy.html__;!... > oVirt Code of Conduct: > https://urldefense.com/v3/__https://www.ovirt.org/community/about/communi... > List Archives: > https://urldefense.com/v3/__https://lists.ovirt.org/archives/list/users@o...

1 0

Re: [External] : Re: can hosted engine deploy use local repository mirrors instead of internet ones?
by iucounu＠gmail.com 24 Jan '24

24 Jan '24

Thanks very for the reply Marcos. I tried another deployment, just to see if "wait for the host to be up" would time out, and I saw a couple of errors in the log: From the ovirt-hosted-engine-setup-ansible-final_clean log, it mentions that the VM IP is undefined: 2024-01-23 12:50:19,554+0000 DEBUG otopi.ovirt_hosted_engine_setup.ansible_utils ansible_utils._process_output:109 {'msg': "The task includes an option with an undefined variable. The error was: 'local_vm_ip' is undefined. 'local_vm_ip' is undefined\n\nThe error appears to be in '/usr/share/ansible/collections/ansible_collections/ovirt/ovirt/roles/hosted_engine_setup/tasks/sync_on_engine_machine.yml': line 2, column 3, but may\nbe elsewhere in the file depending on the exact syntax problem.\n\nThe offending line appears to be:\n\n---\n- name: Set the name for add_host\n ^ here\n", '_ansible_no_log': False} In the ovirt-hosted-engine-setuplog, it mentions not being able to get the storage pool: 2024-01-23 12:50:35,787+0000 DEBUG otopi.ovirt_hosted_engine_setup.ansible_utils ansible_utils._process_output:109 {'changed': True, 'stdout': '', 'stderr': "error: failed to get pool 'localvmvy8whst5'\nerror: Storage pool not found: no storage pool with matching name 'localvmvy8whst5'", 'rc': 1, 'cmd': ['virsh', '-c', 'qemu:///system?authfile=/etc/ovirt-hosted-engine/virsh_auth.conf', 'pool-destroy', 'localvmvy8whst5'], 'start': '2024-01-23 12:50:35.558666', 'end': '2024-01-23 12:50:35.611808', 'delta': '0:00:00.053142', 'msg': 'non-zero return code', 'invocation': {'module_args': {'_raw_params': 'virsh -c qemu:///system?authfile=/etc/ovirt-hosted-engine/virsh_auth.conf pool-destroy localvmvy8whst5', '_uses_shell': False, 'stdin_add_newline': True, 'strip_empty_ends': True, 'argv': None, 'chdir': None, 'executable': None, 'creates': None, 'removes': None, 'stdin': None}}, 'stdout_lines': [], 'stderr_lines': ["error: failed to get pool 'localvmvy8whst5'", "error: Storage pool not fou nd: no storage pool with matching name 'localvmvy8whst5'"], '_ansible_no_log': None} I set the IP to the one I have assigned in DNS, but when I attach to the console of the VM (which is still running, though the disk image has been deleted) via virsh, it shows me a completely different IP: In hosted-engine --deploy, I set a 10.0.0.x address, however, it shows a 192.168.1.x address on the VM. Do I need to set this somewhere else, e.g., with '--ansible-extra-vars=he_ipv4_subnet_prefix='? As for the storage pool, is that for later VM deployment? The deploy script did not ask me for a storage location. If I need to specify this, where do I do this? Thanks again for any help, Kind regards, Cam PS: is there a simple way to have the answers saved so I don't have to keep running through all the questions every time I try a deployment

2 1