1. Home
  2. Xen Orchestra

Subcategories

  • VMs, hosts, pools, networks and all other usual management tasks.

    427 Topics
    3k Posts
    G
    @paco On the stuck server, could you warm migrate the vms to one of the other hosts? Or could you shut down a VM at a time and cold migrate them? The second option is certainly not ideal, but maybe?

  • ACLs, Self-service, Cloud-init, Load balancing...

    97 Topics
    818 Posts
    olivierlambertO
    October release fixed it

  • All XO backup features: full and incremental, replication, mirrors...

    438 Topics
    4k Posts
    P
    Hi, Latest XOA, with fully patched XCP 8.3 here. I'm fiddling around again with NBD+CBT in backup jobs (was avoiding CBT for a time, to reliably control my backups and avoid unnecessary KEY points) in the context of THICK SRs to spare some space. I know that CBT is reset when migrating from one SR to another. But here is what I encounter : VM has no CBT enabled on its VDIs, it is on a SHARED SR in a pool of 3 hosts backup option changed for NBD+CBT, was only NBD before CBT is enabled on the next run by the backup job, I get a delta (was expecting a FULL ?) next run, delta, as expected i migrate this VM on another HOST, without changing its SR CBT is immediatly disabled ? why ?? next run of backup it tries a delta, but "fall back to a full" (normal as CBT has been disabled...), and do a KEY point on the remote next run is a delta as expected does this mean if I do a rolling pool update or host maintenance that will move the all VMs around, all CBT will be disabled and I should expect a FALL BACK TO FULL on all my NBD+CBT enabled backup jobs ??! why disabling CBT on a change of HOST and no move of SR ?

  • Everything related to Xen Orchestra's REST API

    74 Topics
    567 Posts
    olivierlambertO
    But giving your feedback on how you solved it might help many people in the future Thanks!

  • Terraform, Packer or any tool to do IaC

    45 Topics
    424 Posts
    olivierlambertO
    The roadmap depends a lot on the feedback we have on it More demand/popular, faster we'll implement stuff
Log in to post
Load new posts
  • M

    Single backup job two locations

    2
    0 Votes
    2 Posts
    394 Views
    olivierlambertO
    You can have one job with 2 remotes. Or a job with backup and DR at the same time.
  • M

    SAML, automatic user generation

    5
    0 Votes
    5 Posts
    2k Views
    pdoniasP
    @mrbaz Thanks for the detailed feedback! If XO had a way of pulling the user's group information and only allowing those in a certain group to have authorization to XO to have their accounts automatically generated, that would work too. That wouldn't be trivial to implement with our current plugin system. But as you said, a user that logs into XO isn't able to do anything without granting them permissions first. Also, if this is still a concern for you, you can disable user auto-generation in xo-server configuration by adding this line: createUserOnFirstSignin = false However, you'd then have to create the users manually before they can log in. the issue still stands that I can't put a user into a group or assign ACLs until after they have signed in for the first time. I agree but I'm not sure what a good solution to this would be. If we add an option to import all the users at once, it could mean importing a lot of users for some companies, even if many of those users would never actually log in later.
  • D

    Cannot import a large VHD using Import Disk option

    Solved
    8
    0 Votes
    8 Posts
    2k Views
    D
    @nraynaud @olivierlambert Thanks guys!
  • F

    S3 Remote

    12
    0 Votes
    12 Posts
    2k Views
    F
    @lawrencesystems said in S3 Remote: I was hoping to avoid this method as my pfsense is not connected at 10G. Thank you for the clarification. Happy I could help.
  • ForzaF

    un-zstd .xva files exported with XOA?

    Solved
    3
    0 Votes
    3 Posts
    354 Views
    ForzaF
    @olivierlambert said in un-zstd .xva files exported with XOA?: Hi! Again, before starting to talk about technical solution, can you explain the use case? If you have zstd installed on your machine, yes. It's a tar with compression (gzip or zstd). But then it won't be trivial to assemble VM data (but you can exctract metadata easily!) I didn't really mean disassemble the .xva, but rather create an uncompressed .xva that has broader support/portability. The zstd command line tool seems to be able to uncompress it just fine. Thanks.
  • R

    Reset Xen Orchestra Community Edition Web Password

    2
    0 Votes
    2 Posts
    9k Views
    olivierlambertO
    Hey there! It's a good thing to start by reading our doc. If you enter "Password" in the search field, you would have find: https://xen-orchestra.com/docs/troubleshooting.html#recover-web-login-password As you are using Xen Orchestra into production, may I suggest to go for XOA with pro support at some point?
  • E

    Automating VM creation via Terraform

    7
    0 Votes
    7 Posts
    3k Views
    E
    @newknowledge Thanks, I'll try it out later.
  • ForzaF

    Pool Metadata backup sometimes fails when using multiple remotes

    4
    5
    0 Votes
    4 Posts
    384 Views
    ForzaF
    @olivierlambert Maybe. Ive created a support ticket for now. Hopefully you can see if the problem is a configuration mistake on our end or a problem in XO. Thanks.
  • jfrelikJ

    Multiple internal networks in pool for each user

    2
    0 Votes
    2 Posts
    660 Views
    T
    @jfrelik Similar issue here - still trying to figure this out. https://xcp-ng.org/forum/topic/4209/xcp-ng-and-network-virtualisation-help
  • kirovtomeK

    Ansible dynamic inventory and XO

    11
    0 Votes
    11 Posts
    2k Views
    olivierlambertO
    Nice
  • J

    xo-cli - Get alarms

    Solved
    8
    1
    0 Votes
    8 Posts
    880 Views
    olivierlambertO
    You are welcome Enjoy! (great thing about xo cli is that you can get alarms of all your hosts from a single endpoint)
  • X

    XOA migrate the XO itself

    Unsolved
    2
    0 Votes
    2 Posts
    803 Views
    olivierlambertO
    You can't, indeed, live migrate a VM from Intel to AMD (or vice-versa), for obvious reason (the VM will try to call Intel specific instructions after migrating on an AMD CPU, the result would be… a crash!) We had a similar migration ourselves (in our own prod), from Intel to AMD. And we decided to use Continuous Replication to make a first "backup" on destination (the full) while the original VMs were running (no service interruption). Then, we shut down the VMs, make another round of replication (delta, very fast) and then we booted on destination.
  • R

    New VMs created in Xen Orchestra reboot every couple of minutes

    5
    0 Votes
    5 Posts
    780 Views
    R
    @olivierlambert That socket issue was not what was causing the problem, I went through setting by setting and if I choose 2GB of memory for a VM I get constant reboots when using XOA. I was able to repeat the issue with a number of OSs not just RHEL 8 (and variants). I havent went through and compared the differences between XOA and XenCenter yet.
  • ForzaF

    Multiple remotes in a backup - how is retention calculated

    5
    0 Votes
    5 Posts
    875 Views
    badrAZB
    @s-pam said in Multiple remotes in a backup - how is retention calculated: OK Great. So if I set retention to 10, it would keep 10 copies on each remote? Yes. it will keep 10 copies on each remote.
  • F

    "Windows cannot install required files..." Error code: 0x80070570 when trying to install Windows 10 64-bit VM

    1
    1
    0 Votes
    1 Posts
    875 Views
    No one has replied
  • P

    Error: SR_BACKEND_FAILURE_109(, The snapshot chain is too long, )

    6
    0 Votes
    6 Posts
    2k Views
    olivierlambertO
    In theory, Xen Orchestra got a protection against this issue (see https://xen-orchestra.com/docs/backup_troubleshooting.html#vdi-chain-protection ) So you have to find what's causing a long chain outside XO. You can check in XO, in the Advanced tab of the SR view, if there's VDI to coalesce. Or maybe very long chain for other reasons, there's a tool we made called xapi-explore-sr.
  • G

    Backups/VM Migrations Slow

    4
    0 Votes
    4 Posts
    1k Views
    olivierlambertO
    Use CR instead of DR then, it will send only the diff to the other host, so it will be faster.
  • M

    Cannot migrate VM from 7.5/7.6 to 8.2 INTERNAL_ERROR

    4
    0 Votes
    4 Posts
    552 Views
    M
    I can migrate this VM from another host (7.6) to the 8.2 host just fine: [image: 1612363337871-29412a19-7047-4feb-a82e-1b3d3ce61a94-image.png] Im not sure what the agent actually is, thats why im guessing this might be related [image: 1612363441628-eb67e1dd-0979-4b17-94a8-b3d4f72397a5-image.png] VM that fails: [image: 1612363482511-153e71d2-3779-4894-b990-3b4f6b02e032-image.png]
  • ForzaF

    Crashed XOA. Can't boot a clean XOA and use RESTORE because of REGISTRATION

    10
    -1 Votes
    10 Posts
    1k Views
    olivierlambertO
    It's not that easy to block some screens. But that's an interesting suggestion, I'll pass that to @pdonias
  • N

    Multiple VM exports fails with unknown error

    7
    0 Votes
    7 Posts
    2k Views
    DanpD
    @nodesocket Depends on your setup. For mine, I use sudo systemctl restart xo-server.service. You could also just restart the VM.