Search results

Temporary workaround is to boot using old kernel (set default="1>2" in grub.cfg), but it's not fix. Last year I had a similar problem with Lenovo DE storage - the QLogic HBA firmware update helped.

Hi Proxmoxers. After restarting PVE nodes previously upgraded from 5.3.13-1 to 5.4.65-1 kernel, we lost connectivity with LUNs provided by Fujitsu Eternus DX100 S4 - FC interface connected to QLogic 2692 Host Bus Adapters in servers. It's not multipath issue, as we cannot see the storage devices...

Hello, If you start with Proxmox I strongly recommend using the GUI. I suppose that no one here will confirm that the above changes will not break anything. If you want to move VM's harddrive forom one storage to another you should use Move button in hardware section of your virtual machine.

After another Proxmox installation I discovered the strange behavior of the multipath package. When I install this package and NOT reboot server I've got all paths active but hwhandler isn't alua - it has "0" value. When I reboot server one of the paths become faulty as above.

Dear Proxmox'ers, We've got a problem with multipath, because one of the paths is failed: #root@pve:~# multipath -ll 360050763808102622f00000000000006 dm-3 IBM,2145 size=4.9T features='2 queue_if_no_path retain_attached_hw_handler' hwhandler='1 alua' wp=rw |-+- policy='service-time 0' prio=50...

I've just failed (mea culpa) in upgrading from 4.4-1 to 5.2. I followed the instruction from the wiki, however I missed one important (as it turned out to be very important later) of the preconditions - correct repository configuration. I didn't use enterprise repository and also no-subsrciption...

Did you created LVM-thin on your MSA? This storage type can't be shared by hosts/nodes. You cannot add file to LVM-thin storage type because it's block level storage - you have to use file level storage (see: https://pve.proxmox.com/wiki/Storage). Moreover it's not possible to change content...

No, it's not necessary to create LVM group (vgcreate) on the second node, because LVM group is made on external storage which should be visible from two nodes. You can add storage through Proxmox GUI only in the Datacenter section and when added it's propagated to each node belonging to the cluster.

You shouldn't add entries to fstab because there is no need to mount anything. You have to add LVM in Datacenter --> Storage and that's all - Proxmox will take care of serving storage to be used by virtual machines.

First of all you have to define second server and mappings in MSA configuration. Since storage devices are defined at Datacenter there is no need to add storage on each node/server. If you want share LVM between servers you have to set up Proxmox Cluster and add second server to cluster.

Congratulations Zaqen. It is worth to be patience. I hope that usage of PVE rewards all your suffering ;) The problem was in size of volume made on MSA as I suggested, am I right? Proxmox has nothing to do with support of GPT or LVM because it's based on Debian with Ubuntu kernel. I checked out...

Hallo Zaqen, It's very strange. Are you sure that definition of hosts and hosts mappings in MSA are correct? Maybe they are read only mappings or the volume's setting at MSA does not allow to write on it down? I would suspect configuration of the SAN switches - zoning?. I think that is not the...

no What about requirements?

Yes, it's (sometimes) normal behaviour. Zaqen, could you check output of this command: pvcreate -vvvv /dev/mapper/3600c0ff0001ae601d4d8b05a01000000 If you will find something like "Skipping: Unrecognised LVM device type 253" you should add found magic number to /etc/lvm/lvm.conf (in the...

I don't think that Zaqen have to configure this, because of native Linux multipath is now hardcoded to Linux. Look at https://support.hpe.com/hpsc/doc/public/display?docId=mmr_kc-0133803

Please, show actual output of multipath -ll command

Don't even think about it! It's not the way out.

Maybe, the problem is that you're using wwids in binding file that are not present in the system. Remove any records from /etc/multipath/bindings file and remove /etc/multipath.conf file. Reload multipath daemon using multipath -r command. In the output of multipath -ll command you will see...

Files in /etc/multipath/ directory must have the same wwids inside! Syntax in /etc/multipath/wwids is ok, but you should not to edit this file.

It seems that you've got friendly names because you're using multipath configuration files: /etc/multipath.conf or/and /etc/multipath/bindings I'm not sure but it can be some issues creating LVM on GPT disk or some issues with LVM filtering. You could try to remove partitions from disk and...