I have isolated the problem. The issue was with dynamic memory allocation on our test server. This was definitely a localized issue. Cheers

@olivierlambert https://docs.microsoft.com/en-us/windows-server/remote/remote-desktop-services/clients/remote-desktop-uri a plain rpd://ip-address seems to work fine in Mac OS and iOS but more information may be needed on windows clients not sure if the uris work on all systems the same. EDIT Looks like windows doesn't support URI to RDP but Microsoft makes it work for basically every other OS. I still think its a worthwhile feature even if it is slightly less functional because Microsoft is weird.

Congratulations!

@lcy985a nice one. I ran 19.10 in this morning as I need Wazuh setup pretty sharpish for an an audit.

Normal backup jobs. In this case it is the "full" copy of the VDI I just moved to a local SR (due to maintenance on the NFS) I mentioned here https://xcp-ng.org/forum/topic/2247/moved-vdi-delta-backup-restarts-on-full. I see that Control Domain load is high and in top I see this: At the moment it is only one disk left to transfer, but there were 3 to start with, so the load was higher in the beginning.

OH ok. Hope you can prioritize this soon

Haven't experienced that issue. Might help if you provided additional details, such as -- When do you last pull from source? What type of SR(s) should be showing (Local, remote, etc)? Are you able to select a target SR when defining a new backup job? Have you opened the Dev Tools (F12) and checked the console for any errors? Ditto for the system logs? FWIW, I know the restore functionality was broken by a recent commit (different issue than you listed), but that appears to have been fixed by yesterday's commit. If you haven't recently updated, then I would suggest that you take a snapshot of your VM, update to the latest source, and then retest your restore operation.

kids aren't alright

@tuxen I never saw such "performance mode" only very basic CPU settings, like cstate. When i had C-state to auto I had never any issues with reboots (only had it running for 2 days straight), before I disabled the C-state. I found out having C-state to auto my processors never reached their max clock frequence. I did however never test to change the scaling-governor to performance. I did not need to change this to performance to get full clockspeed when C-state was disabled. I just had to load the CPU.

Your CR issue might be related to a XO bug on master, not too a XCP-ng issue

Sorry, I was trying to execute the git command in the wrong directory, my bad. Everything worked perfectly, the restores worked now. Thank you very much @Danp.

I'm working on it at this very moment, I'll let you know what I find.

@petr-bena said in XCP 8.1 with Kernel 5.3 : @maxcuttins current kernel supports RBD, which feature are you missing? I am using kernel RBD and it works OK "Exclusive Lock" is not included in the 4.19 kernel. So it is for "object-map" and "fast-diff".

@pdonias We only see the connection error between xo-server and xo-updater, we need to reproduce to figure out what's going on.

Hi, There is different things one must consider during CR: SR type: block based or file based? First one will be thick provisioned, so plan a lot of space (eg: local LVM, LVM over iSCSI etc.). Local ext/XFS, NFS or XOSAN aren't affected by this problem Snapshot left: to do a delta, we must have a way to compare to something. The snapshot which left on the SR is the one that will be used for the next backup. It will be compared to the current disk state and only the delta will be sent. So that's perfectly normal "Job canceled to protect the VDI chain": this is a protection to avoid doing another backup while the previous removed snapshot isn't coalesced. You have a link in the message, please refer to the documentation for details

@KFC-Netearth Welcome