Recent content by mgabriel

Was steht denn im Task Log für eine Fehlermeldung, wenn du eine der VMs/CTs startest?

Was geschieht denn, wenn du den Befehl /sbin/lvs --separator : --noheadings --units b --unbuffered --nosuffix --options lv_size <volume_group> von Hand eingibst, einmal wenn keine Backups laufen und einmal wenn Backups laufen? Wie hoch steht denn der iowait zur Zeit des Backups auf dieser LUN?

Der qm guest exec verwendet auch die API :). Geht also auch direkt per API mit https://pve.proxmox.com/pve-docs/api-viewer/index.html#/nodes/{node}/qemu/{vmid}/agent/exec

Proxmox VE has the ZED utility already installed and sends mail to root in case of a ZFS pool degradation. You can change the notification to an email address you like [1]. [1] https://pve.proxmox.com/pve-docs/pve-admin-guide.html#chapter_zfs -> 3.9.7. Configure E-Mail Notification with ZFS on...

Wenn es sporadisch auftritt, könnte es sein dass eine deiner Storages zu langsam ist. Läuft dein PBS virtuell auf den gleichen LUNs? Das konnte ich dem ersten Satz nicht so richtig entnehmen.

It depends on what your DC supports. I guess the only possibly usable solution for your scenario will probably be QinQ. If you want to use your private VLAN as the outer/service VLAN. Then you'll be able to use any VLAN you want inside the outer service VLAN [1]. Please keep in mind that the...

Yes, but ceph status also shows how many other components there are. I guess you have a lot of OSDs, but I can only guess until you send the output of ceph status. Ah, and if you like, you could also send the output of ceph health detail.

If you just want to be able to start your VMs/CTs on the remaining node, you can issue a pvecm expected 1 on the commandline of the remaining node to reduce the number of votes to 1 and regain quorum.

could you post the output of ceph status?

Awesome :) Congratulations, this is huge.

@martin2089 danke dass du deine Lösung hier noch gepostet hast. Die hat mich hier her geführt :-). Anzeigen kannst du den Status übrigens mit `-d`, siehe Beispiel: bridge -d link show [dev DEV] 12: tap181i0: <BROADCAST,MULTICAST,PROMISC,UP,LOWER_UP> mtu 1500 master fwbr181i0 state forwarding...

Congratulations for another fine release :-) Proxmox rocks.

According to the logs, the Disk "SauvegardeUSB" is still failing: Nov 09 11:15:48 pve pvestatd[2853]: could not activate storage 'SauvegardeUSB', zfs error: cannot import 'SauvegardeUSB': I/O error

Your 42 GB image is cut into 4 MB chunks. Those chunks are placed under the .chunks path of your datastore (like /mnt/datastore/mydatastore/.chunks). To be able to restore those chunks into your original 42 GB image, you need an index of all the chunks - and that's your drive-ide0.img.fidx (fidx...

Is there something in the syslog / journal log at that time that gives some hints what happens? What do you mean by "complete stop" of the cluster? What happened?