Posts made by karlisi

@Ascar In XO cloud means pool, "hamburger" means host. If VM is running, it belongs to host where it runs, and receives host icon, no matter where it is kept in stopped state. If VM is stopped, it belongs to pool if kept on pool shared storage (cloud icon), or to host, if on host's local storage (hamburger).

Sometimes we want to try harder
Perhaps the confusing part is, if I go to https://xen-orchestra.com/ there is no download link. One could not realize, the download option is behind login. I'm only guessing.

@olivierlambert I believe, it is from here
https://xen-orchestra.com/docs/xoa.html#via-a-manual-xva-download

Installed AD DS role on my test Winserv2025 VM and conhost processes are appeared and already gone from 7 to 10 in about hour. In the same pool Winserv2022 AD DC VM no conhost processes at all.

WinServ 2025 on xcp-ng 8.2.1, xcp-ng tools v.8.2.2 rc1. No additional roles, except Windows Admin Center preview installed. Uptime about 2 months. No conhost processes at all.

If you need to restore one DC from VM backup, this can help
https://learn.microsoft.com/en-us/windows-server/identity/ad-ds/manage/virtual-dc/restore-virtualized-domain-controller#restore-the-virtual-dc-with-a-vhd-file

@waveguide said in What are recommended ways to grant XCP-ng and XO access to the internet?:

Do you poke holes in your firewall so the Management network can resolve DNS requests and reach the IP addresses

Outgoing communication from management and xcp-ng hosts to Internet is not very dangerous, because hosts are the initiators. Generally I don't block it. You can monitor these connections or restrict to certain addresses.

I am using RSS reader, feed https://xcp-ng.org/forum/recent.rss
Not the most elegant three click solution to read new replies in watched topics, click on unread posts (open envelope icon), then on All topics button, it opens menu, there click on Watched topics.

@chaoticmaster I suggest to use xcp-ng tools
https://xcp-ng.org/docs/guests.html#citrix-tools-vs-xcp-ng-tools

ISO works, installed already on 2 Ubuntu 22.04 VMs

@planedrop To avoid confusion, XOA is paid product, ready to use virtual application. If you installed XO from sources, it's simply XO or XOCE (XO Community Edition). XO components versions you can see in About menu, both xo-server and xo-web. If you want to know how it corresponds to XOA version, see changelog https://github.com/vatesfr/xen-orchestra/blob/master/CHANGELOG.md

@geoffbland said in Xen Orchestra Application XOA Deployment Options/Best Practices.:

What exactly do you mean when you say "XCP-ng small web UI" ? I don't think I have heard of that before?

If you go to http://your.xcp-ng.host.ip you will receive web UI with some links to documentation and to management tools, including Xen Orchestra Virtual Appliance quick deploy wizard

XOA is only the version we (Vates) distribute, not the version from the sources

What is the correct name for the version from the sources then? Is that just called Xen Orchestra?

They are calling community version (from sources) XO and precompiled virtual appliance (paid version) XOA.

@julien-f said in CR tag ignored (kind of) in smart backup:

incomplete VMs should be ignored by smart mode

I believe, XO is using some other way to prevent backup loop.

Thanks again, now CR works as expected!
xo-server 5.64.0
xo-web 5.66.0

@julien-f Thanks!

XO CE, xo-server 5.62.0, xo-web 5.63.0 (doesn't matter, the same was on older versions too)
We have 2 pools (one host XS 7.0.0 in each), cross-pool backups using continuous replication configured on both of them. It means, VMs from pool 1 are replicated to pool 2 and VMs from pool 2 are replicated to pool 1. To avoid replication back, causing loop, we are using exclusions by tag 'Continuous Replication' (default tag XO adds to every replica). Both backup jobs are overlapping in time.
The problem is, some replicas are added to target's backup job and replicating back, despite having 'Continuous Replication' tag. I suspect, that tag is added only after replication is completed, and incomplete VMs are without this tag. This can explain why they are added to backup job. Am I rigt? If so, how can I avoid this? Perhaps this can be considered as XO bug and patched?

@x-rayd said in Backup Continuous Replication hängt!:

This problem have only pool-master node, other node does not have the problem.

Perhaps. I have 2 pools with one node each.

I have the same problem since... I don't remember, a year perhaps. Orphaned snapshots are some way related to continuous replication, they appears randomly sometimes on almost all VMs, sometimes none. For me this is not a problem, I delete them as part of daily routine.
I am using XOCE, xo-server and xo-web 5.59.0.

@tts Are You on latest XOA version?

Stuck CR is a problem with latest XO versions and XO team is working on it.
You can delete incomplete VMs [Importing...] and you should restart XO before start CR jobs again. As workaround some users here made cron jobs to restart XO daily to avoid CR failures.
Automatically deleting incomplete VMs can be dangerous, IMHO. Some of these [Importing..] VMs are properly replicated, I examined few Windows Servers and they are starting and working. So, they can be used as backups.