florent Thanks for the Tip!

The risk of an infinite delta and losing the backup chain if something goes wrong is a bit too high I bet if something goes wrong exactly in that moment, we need a reliable backup chain

Sadly, we do not have "cheaper" storage for mirror backup, as it's only a small setup for a small company.

This situation is a bit of a pickle if you want one full backup on a certain day, plus deltas, but not two full backups all the time

Is there any chance that this will change in the future to make that situation happen?

Danp very interesting! Thank you! So is it somehow possible that we only do full-backups on Saturday night and only have one fullbackup? Our backup storage is almost full, and we need to make some space, so getting rid of double full backups would be great cheers Ringo

Hi everyone

we have a strange behaviour at the moment, we have a delta backup job configured like that:

• Full backup interval is 0 as we want full backups only on Saturday night
• We have one schedule for Sunday to Friday night with no forced full backup for incremental backups
• We have one schedule for Saturday with Force Fullbackup enabled to make sure only Saturday is the full backup

Screenshot 2025-04-15 at 21.29.47.png

So, now problem is we have two full backups on our remote, and we are getting into storage space trouble

Screenshot 2025-04-15 at 15.11.58.png

Has anybody an idea why this is happening? What do I miss?

cheers and have a great day!
Ringo

astrugatch you can put it into the same files like the cert it’s called a chain cert then

julien-f thanks! I have tested it and its working now! cheers Ringo

julien-f happy to help let me know when you fixed it and I will test it. cheers Ringo

hi everyone,

I am on the last version of XO and wanted to play around with the REST-AP. scheduling a rolling pool update but I got an error I can't figure out:

curl -X POST -b authenticationToken=XXX https://xo.xxx/rest/v0/pools/5222b44b-e25b-a14d-4da2-c7a0dd8cf59f/actions/rolling_update
/rest/v0/tasks/0m1agafzh

the log says:

curl -b authenticationToken=XXX https://xo.xxx/rest/v0/tasks/0m1agafzh
{
  "id": "0m1agafzh",
  "properties": {
    "name": "REST: rolling_update pool",
    "objectId": "5222b44b-e25b-a14d-4da2-c7a0dd8cf59f"
  },
  "start": 1726820415341,
  "status": "failure",
  "updatedAt": 1726820415449,
  "end": 1726820415449,
  "result": {
    "message": "Cannot read properties of undefined (reading 'user')",
    "name": "TypeError",
    "stack": "TypeError: Cannot read properties of undefined (reading 'user')\n    at file:///opt/xo/xo-builds/xen-orchestra-202409201003/packages/xo-server/src/xo-mixins/xen-servers.mjs:704:82\n    at AsyncLocalStorage.run (node:async_hooks:346:14)\n    at Task.runInside (/opt/xo/xo-builds/xen-orchestra-202409201003/@vates/task/index.js:169:41)\n    at Task.run (/opt/xo/xo-builds/xen-orchestra-202409201003/@vates/task/index.js:153:31)\n    at XenServers.rollingPoolUpdate (file:///opt/xo/xo-builds/xen-orchestra-202409201003/packages/xo-server/src/xo-mixins/xen-servers.mjs:703:16)\n    at rolling_update (file:///opt/xo/xo-builds/xen-orchestra-202409201003/packages/xo-server/src/xo-mixins/rest-api.mjs:634:11)\n    at Task.runInside (/opt/xo/xo-builds/xen-orchestra-202409201003/@vates/task/index.js:169:22)\n    at Task.run (/opt/xo/xo-builds/xen-orchestra-202409201003/@vates/task/index.js:153:20)"
  }
}

Do you have any idea what the problem is here?

If I start the update via XO everything works without errors.

thanks and have a great day!
cheers Ringo

Studmuffn1134 No we shutdown virtual machines first and xcp-ng servers last.
Puppet is a config-mgmt tool so you don't have to install apcupsd on every machine manually, it can automates stuff like that.
what do you use as nfs storage? normally you can configure to a shared network storage device that it talks to the smart-ups and shuts it self down before its running out of power

cheers Ringo

We installed apcupsd in every virtual machine with puppet and shut down test/dev systems early and important systems later. XCP-ng servers also got it installed and shutdown last.

jedimarcus thank you i will have a look!

hi team,

we were wondering, is it possible to get the usage of the SRs of a pool from the xoa API so we could monitor it?
also is there a list somewhere online of what information we can get from the xoa API at the moment? I don't want to bother you with my questions all the time, easier to check what's possible

cheers, Ringo

julien-f olivierlambert thanks guys I will wait until the end of the month you rock!

hi team, thanks for your excellent work! we love xcp-ng and xoa
I've written an ansible playbook to update our virtual machines and I would like to trigger a snapshot of the VM bevor I start the automatic update. is there a way to trigger that snapshot through the rest API? I've searched a bit in documentation and the forum but couldn't find something
cheers Ringo

Thanks guys! Absolut Legends

julien-f thanks for the answer! this will get overwritten with every xo source update right? how about a licensed xoa? cheers Ringo

Hi Team,
i want to use the xoa api to monitor if i have old or too many snapshots connected to an VM, thank you for implementing the feature to your API that it is possible now. I check it via icinga and need an token which expires in 5 years or something and not after a month. when I try to create a user in the webfrontend admin user with a date 5 years or 11 month from now, it logs an error with:

too high expiresIn duration: 1956120152841

I also tried to create the user via cli:

#create timestamp for 5 years from now:
date +%s -d "5 year"
1831359954

#create user token
xo-cli --createToken --allowUnauthorized --expiresIn 1831359954 localhost admin@admin.net

Authentication token created

7kkiyVwzgISlOTO2LozsjtVTDPHG71lvqHEtYeSRQf0

but the expiration date of the token is only 30 days from now.

I dont want to create a new token every 30 days and reconfigure our icinga monitoring system to work with a new token, so can you please tell me how I can create a tekoen with higher age and maybe put it into the documentation, i was able to find the "--expiresIn" parameter but no info what you should put in there

cheers guys! Ringo

Steve_Sibilia hey, what icinga check are you using for snapshots? do you have any xcp-ng related icinga checks running?

julien-f that also fixed my issue, cheers!

Hi everyone, I am looking for a way to find snapshots in my pool that are pretty old so I can remove them. finding the "type" snapshot is not complicated but seeing the creation date and sorting them by that is my actual problem, is there a way for that?
cheers Ringo

fohdeesha thanks for the information, I've added the following to my cloud-init template:

runcmd:
  # disable dhcp for eth0
  - [ sh, -c, sed -e '/iface eth0 inet dhcp/s/^/#/g' -i /etc/network/interfaces ]

at the end of the installation I have a reboot of the machine anyway so I don't restart the network in between:

power_state:
 delay: "+2" # minutes
 mode: reboot
 message: rebooting system after cloud-init configurations are finished
 timeout: 120 # seconds

cheers and keep up the good work!
Ringo