- Users - oVirt List Archives

Could not read L1 table: Input/output error (libgfapi enabled)
by Николаев Алексей 13 Oct '20

13 Oct '20

1 0

Imageio Daemon not listening on port 54323
by tim-nospam＠bordemann.com 13 Oct '20

13 Oct '20

Hello. After an upgrade I am not able to upload images anymore via the ovirt ui. When testing the connection, I always get the error message "Connection to ovirt-imageio-proxy service has failed. Make sure the service is installed, configured, and ovirt-engine certificate is registered as a valid CA in the browser.". I found out that the imageio daemon doesn't listen on port 54323 anymore, so the browser can not connect to it. The daemon is configured to listen on port 54323 though: # cat /etc/ovirt-imageio/conf.d/50-engine.conf [...] [remote] port = 54323 [...] The imageio daemon has been started successfully on the engine host as well as on the other hosts. I am currently stuck, what should I do next? The ovirt version I am using is 4.4. There is one machine running the ovirt engine and there are 2 additional hosts. The OS on the machines is Centos 8. Thank you, Tim

4 7

Upgrade oVirt Host from 4.4.0 to 4.4.2 fails
by Erez Zarum 13 Oct '20

13 Oct '20

Hey, Bunch of hosts installed from oVirt Node Image, i have upgraded the self-hosted Engine successfully. I have ran Check Upgrade on one of the hosts and it was entitled for an upgrade. I use the UI to let it Upgrade, after multiple retries it always fails on "Prepare NGN host for upgrade." so i chose another host as a test. I have set the Host into Maintenance and let all the VMs migrate successfully. made sure i do have the latest 4.4.2 repo (it was 4.4.0) yum install https://resources.ovirt.org/pub/yum-repo/ovirt-release44.rpm And issued "dnf upgrade" Installing: ovirt-openvswitch replacing openvswitch.x86_64 2.11.1-5.el8 ovirt-openvswitch-ovn replacing ovn.x86_64 2.11.1-5.el8 ovirt-openvswitch-ovn-host replacing ovn-host.x86_64 2.11.1-5.el8 ovirt-python-openvswitch replacing python3-openvswitch.x86_64 2.11.1-5.el8 Upgrading: ovirt-node-ng-image-update-placeholder Installing dependencies: openvswitch2.11 ovirt-openvswitch-ovn-common ovn2.11 ovn2.11-host python3-openvswitch2.11 Installing weak dependencies: network-scripts-openvswitch network-scripts-openvswitch2.11 It was very quick, but nothing else happened, I did try to reboot the host but I still see the host as oVirt 4.4.0 and as expected it still says that an update is available.

5 7

Ovirt 4.4/ Centos 8 issue with nfs?
by lee.hanel＠gmail.com 13 Oct '20

13 Oct '20

Greetings, I'm trying to upgrade from 4.3 to 4.4. When trying to mount the original nfs items, I'm getting the following error: vdsm.storage.exception.StorageServerAccessPermissionError: Permission settings on the specified path do not allow access to the storage. Verify permission settings on the specified storage path.: 'path = /rhev/data-center/mnt/nfshost:nfs_path' with the following stack trace: 2020-10-08 19:00:17,961+0000 ERROR (jsonrpc/4) [storage.HSM] Could not connect to storageServer (hsm:2421) Traceback (most recent call last): File "/usr/lib/python3.6/site-packages/vdsm/storage/fileSD.py", line 82, in validateDirAccess getProcPool().fileUtils.validateAccess(dirPath) File "/usr/lib/python3.6/site-packages/vdsm/storage/outOfProcess.py", line 194, in validateAccess raise OSError(errno.EACCES, os.strerror(errno.EACCES)) PermissionError: [Errno 13] Permission denied During handling of the above exception, another exception occurred: Traceback (most recent call last): File "/usr/lib/python3.6/site-packages/vdsm/storage/hsm.py", line 2418, in connectStorageServer conObj.connect() File "/usr/lib/python3.6/site-packages/vdsm/storage/storageServer.py", line 449, in connect return self._mountCon.connect() File "/usr/lib/python3.6/site-packages/vdsm/storage/storageServer.py", line 190, in connect six.reraise(t, v, tb) File "/usr/lib/python3.6/site-packages/six.py", line 703, in reraise raise value File "/usr/lib/python3.6/site-packages/vdsm/storage/storageServer.py", line 183, in connect self.getMountObj().getRecord().fs_file) File "/usr/lib/python3.6/site-packages/vdsm/storage/fileSD.py", line 93, in validateDirAccess raise se.StorageServerAccessPermissionError(dirPath) via an ls, it looks like there are the correct permissions: ls -alh total 0 drwxr-xr-x. 1 vdsm kvm 100 Oct 8 19:15 . drwxr-xr-x. 4 vdsm kvm 115 Oct 8 19:02 .. drwxr-xr-x. 1 vdsm kvm 52 Oct 1 20:35 ffe7b7bb-a391-42a9-9bae-480807509778 d---------. 1 vdsm kvm 22 Mar 17 2020 '#recycle' But, I wrote a short script to check the individual permissions that are checked, and it thinks I can't write to the directory: We can Read From it: /rhev/data-center/mnt/nfshost:nfspath We can't write to it: /rhev/data-center/mnt/nfshost:nfspath We can execute to it: /rhev/data-center/mnt/nfshost:nfspath This is doing a simple: print(os.stat(path)) if os.access(path, os.F_OK): print("It Exists", "/rhev/data-center/mnt/nfshost:nfspath") else: print("It Doesn't Exists", "/rhev/data-center/mnt/nfshost:nfspath") if os.access(path, os.R_OK): print("We can Read From it: ", "/rhev/data-center/mnt/nfshost:nfspath") else: print("We can't Read from it: ", "/rhev/data-center/mnt/nfshost:nfspath") if os.access(path, os.W_OK): print("We can write to it: ", "/rhev/data-center/mnt/nfshost:nfspath") else: print("We can't write to it: ", "/rhev/data-center/mnt/nfshost:nfspath") if os.access(path, os.X_OK): print("We can execute to it: ", "/rhev/data-center/mnt/nfshost:nfspath") else: print("We can can't to it: ", "/rhev/data-center/mnt/nfshost:nfspath") I took this same checks over to the Centos 7 host, and I it passes the checks fine. Does anyone have any ideas? Thanks, Lee

4 7

VM reboot cause SERVER reboot (oVirt 4.4.2)
by luigi.voso＠convergenze.it 12 Oct '20

12 Oct '20

Hi all, I've a SUPERMICRO server with a M10 GPU (at the moment i've a free-trial license) and i'm using oVirt 4.4.2. My issue is very strange and it seems to be a bug because i can reproduce it: 1. Create a VM (i've tested on Win10Pro/WinServer2012R2) with a vGPU and install the driver (without driver installed, this issue does not appear); 2. Restart the VM. At this point the physical server restarts. Instead, when i shutdown the windows vm, this issue is not appear. I've also tried with a centos 8 vm and it works very well (no reboot). All NVIDIA driver's are update at the 2020-09-30 latest release (11.0). All windows driver are installed (virtio-win-0.1.189). The VM works very well and the vGPU too. This is my very first test with vGPU and i can't understand the reason of this issue. NVIDIA enterprise support told me that it's an ovirt problem, because Windows guest operating system are supported only under Red Hat subscription programs. Has anyone else had the same problem or can you help me? Thank you in advance

1 0

Last week for participating to oVirt Survey Autumn 2020
by Sandro Bonazzola 12 Oct '20

12 Oct '20

Hi, just a kind reminder this is last week for participating to oVirt Survey Autumn 2020! The survey is closing on October 18 2020. You can submit your response at https://docs.google.com/forms/u/1/d/e/1FAIpQLSdzzh_MSsSq-LSQLauJzuaHC0Va1ba… *Please note the answers to this survey will be publicly accessible.* This survey is under oVirt Privacy Policy available at https://www.ovirt.org/privacy-policy.html -- Sandro Bonazzola MANAGER, SOFTWARE ENGINEERING, EMEA R&D RHV Red Hat EMEA <https://www.redhat.com/> sbonazzo(a)redhat.com <https://www.redhat.com/> *Red Hat respects your work life balance. Therefore there is no need to answer this email out of your office hours. <https://mojo.redhat.com/docs/DOC-1199578>* * <https://www.redhat.com/it/forums/emea/italy-track>*

1 0

oVirt 4.4 - HA Proxy Fencing
by Jeremey Wise 11 Oct '20

11 Oct '20

I have a pair of nodes which service DNS / NTP / FTP / AD /Kerberos / IPLB etc.. ns01, ns02 These two "infrastructure VMs have HA Proxy and pacemaker and I have set to have "HA" within ovirt and node affinity. But.. within HAProxy, the nodes use to be able to call the STONITH function of KVM to reset a node if / as needed. Fence Device: "stonith:fence_virsh" But with oVirt this function no longer works. For VMs and other applications which need STONITH functions, how would this be done with oVirt. I would assume the call would be to oVirt-Engine now, but not seeing documentation on HA-Proxy / Pacemaker to use oVirt. Can someone point me in the right direction? -- penguinpages

2 1

Fencing with oVirt 4.4 apc master switch
by Jeremey Wise 11 Oct '20

11 Oct '20

I use to have hand built CentOS+KVM + Gluster. Moved to HCI oVirt controlled system and so far, been very happy with stability and quality. One feature that was working in old "hand build" version was fencing. I have an old APC master switch AP9606. I use to have and use it for master switch functions of nodes but sense replacement with oVirt it has not worked. Here are setup notes and symptoms IP: 172.16.100.2 Username: apc Password: <blah> Port 8 (it is an eight port unit) Slot: <empty as I have no idea what is expects here I also input values here but no change> Error: Test failed: WARNING:root:Parse error: Ignoring option 'agent' because it does not have value ERROR:root:Unable to connect/login to fencing device 2020-10-11 11:28:50,885 ERROR: Unable to connect/login to fencing device Questions: 1) Has anyone setup this with oVirt such that it uses apc master switch 2) Where is oVirt putting these logs / running this service. I assumed oVirt engine running the services and so tried to find details and logs here. -- penguinpages

1 0

ovirt broker how to modify network_test dns ip address?
by wodel youchi 11 Oct '20

11 Oct '20

Hi, How to modify the ip address of the DNS server used by ovirt-broker for testing? Regards.

2 1

too many glitches while moving storage domain
by Jiří Sléžka 11 Oct '20

11 Oct '20

Hi, today I started one not so common operation and I would like to share my (not so good) experience. I have 4 old Opteron G3 hosts which creates 1 DC with 1 cluster with 4.2 compatibility. Then I got 2 newer Intel based hosts which creates separate DC and cluster. I use one shared FC storage with few LUNs for all the stuff. Intel cluster is selfhosted with oVirt 4.4.2 where I migrated original standalone oVirt 4.3.10 manager. So I have 2 DCs, one with 4.2 and one with 4.4 compatibility. Now the funny part... I had 2 LUNs connected to old 4.2 DC. My intention was detach one of them and attach it to the new DC. first pain was that some of vms had ther disks on both LUNs. It is indicated late in the process and without hint which vms they are. So I had to activate LUN in old DC and tried to find that vms (search string in vms tab "Storage = one and Storage = two" seems not working). Ok, it took two or three rounds... then, also late in the process there was problem that one vm had previewed snapshot so another round with activation of LUN in old DC... Then I was able detach and import LUN to new DC. Nice, but with warning that LUN has 4.2 compatibility which will be converted to 4.4 and there is no way back to connect it to old DC... It is logical but very scary if something went wrong... but it did not in my case :-) LUN is connected in new DC. Now I had to import vms. Most vms were imported ok but two of them were based on template wich resides on other LUN. It was not indicated during detaching! It looks like I cannot move template from storage to storage other way than through Export storage (which I don't have at this moment) or through OVA export for which I have not enough free storage space on hosts. Its a trap! :-) Btw. there is no check for free space while starting export to OVA (template uses preallocated disk). Exporting task still runs but there is no free space at the host... and probably no way to cancel it from manager :-( Ok, I had most of the vms imported. Last really strange thing is that I lost one vm during import. It is not listed in VirtualMachines nor in VM import tab on starage nor in Disk tab... that vm was an Aruba migration tool and was imported from OVN image. In fact there are two disks in Disk import tab, one of them has no Alias/description and was created today around time I started work on this migration. The second one has alias "vmdisk1" and is few months older but I have no idea if it is the lost vm... Sorry for long story, TL;DR version could be: There are glitches in some (not so common) workflows... Cheers, Jiri

2 1