Recent content by fips

Thanks to your support I could get it done: if once someone need a hint DATE=$(date '+%Y-%m-%d' -d "1 day ago") FILEall=$(pvesm list my.pbs --vmid 810 | grep ${DATE}) FILE=$(echo ${FILEall} | awk '{print $1}') qmrestore ${FILE} 810 -storage HDD_36T

hmm well this would work somehow if the pbs are mounted, but out of the box the pbs storage isn't. Well I don't want to harm the backup infrastructure doing some hacky stuff..

Hey Guys, I am thinking about a scheduled (via Cron?) automatic restore of VMs which have been backed up via PBS. My first attempt was via CLI but I can set the date via variable but not the timestamp as wildcard. qmrestore my.pbs:backup/vm/710/2025-04-18* 810 -storage HDD_36T unable to parse...

Because I have to set "pvecm expected" on my own through shell or console after nodes have restarted. I need a setup where a specific VM can be startet by pve in any case. My idea was through a config setting in the corresponding .conf file of the VM.

Hello, I am running a 2 node cluster with an external QDevice. So I am looking for a solution to force start a VM even just 1 node is online and the QDevice is not reachable. That VM is not running on a shared storage so I don't care about split brain. I don't want to set expected votes or...

I guess I understood it right but let me ask different: Transfer last syncs all chunks even from the very beginning of the backup retention which are needed to be able to create the last snapshot?

Do I understand right, that "transfer last" still syncs the first full backup ever made and all increment data to perform a full restore of each N snapshot on the destination side?

How can I achieve to run backups to an "odd" storage on a odd week and to an "even" storage" on an even week? I have tried with the simulator but it was unsuccessful

Ich merke gerade, dass ich das nicht verstanden habe Ich dachte immer das auch bei ZFS die Art des RAID Levels entscheidet wie viele Disks für Parität verwendet werden und daher eine gewisse Anzahl an Laufwerksgrößen von der Bruttokapazität abgezogen werden muss. Das nun auch das volblocksize...

Gute Glaskugel! Habe wirklich auf dem 2. Node ein Z1 mit 6x800G. Dachte er nimmt die Blocksize wie bei NodeA, zumindest stand der gleiche Wert (8k) (und thin provisioning) bei Storage. Oder wo stelle ich Volblocksize ein? Danke!

Hallo Leute, habe auf NodeA ein ZFS Mirror mit 3,71TB Storage, davon 2,32TB in use (62,59%). Nachdem ich heute Replikation auf NodeB eingerichtet habe mit einem 3,71TB Storage ist dieses schon zu 3,62TB voll (97,68%). Wie kann das sein?!? Müsste der Speicherplatz auf NodeB nicht der gleiche...

Hallo Leute, habe einen Nested HyperV Replication Server den ich gerne zuverlässig sichern möchte. Mein erster Plan war die VM einfach via PBS zu sichern, habe aber bedenken wegen der laufenden Replikation vom anderen HyperV Server. Ich fürchte inkonsistente Daten, da sich ja alle paar Minuten...

Ah, yes Can I somehow upvote? +1

I am sorry, but the comment is not shown, see attachment. When I open the specific backup (by clicking on + at for eg. vm/101), I can do see the comments but not at root level of each backup.

What about to leave the user the possibility to choose if the VM/CT name is the default note? I would definitely activate that option. Especially with the remote sync jobs its nice to read in clear names what is that backup from.