FYI, I have just published security updates today PLUS the fixed ixgbe driver as an official update to XCP-ng 8.1 and 8.2.

We made it. This is the end of this huge thread.

A big thank you to everyone involved in debugging the issue.

And this is not a 🐟 :D.

When you boot in UEFI mode, press "e" to edit the boot command line, you have a line to change the memory for dom0.

No, I don't know where the problem resides exactly.

I did the manual installation via the XVA. It worked for me.

@akurzawa said in Extrlemy slow backup speec = about few MB/s:

@nikade said in Extrlemy slow backup speec = about few MB/s:

Maybe you saw the same thing I did - I also had major issues with backups. Then in the summer of 2019 I upgraded our XOA (Yes, we have the paid version) and everything was going 3-4x times faster. Must have been some major improvement in the XOA code.

So maybe some official info that if you have XO from source Your backup speeds will be as low as 3-4 MB/s? Or what?

Sorry if I was unclear - XOA is built from the same sources so there is no difference in that aspect.
All tho, there was a big performance issue regarding backups which was resolved.

I am using XO from sources in my "private" infrastructure and I can assure you there are no "limits" on purpose.

How about doing your own matrix?

I'm not sure Unitrends is supporting XCP-ng. Please refer to your backup vendor.

Current backup solutions claming to support XCP-ng: https://xcp-ng.org/docs/backup.html

@chrisarzu

You're very welcome. My solution to the similar problem I'd had was to set up a couple of internal systems as NTP servers so that I always had something with the right time and static IP addresses and pointed everything needing NTP at them.

@dvdhngs

install onto (mdadm/ZFS) RAID 10: no

to use (mdadm/ZFS) RAID 10 as SR: yes, of course!

P.S.:

I gave my latest "trick" another run:
Did a full copy of the VM within xo to the other host. While that job ran, I started the delta backup that finished OK. After the copy was done I deleted the copy and saw - as last time - that the host was coalescing. After finishing the sr's advanced tab is empty again and stayed empty.

@jokurinopoloski I reviewed your PR!

Thanks for the contribution 🙂

So your post was great and its seriousness provided traction to move forward. Congrats!

Andy is one of the best (likely the best) Xen expert I know. He's also a bit "pessimistic" sometimes, one would say realistic 😛 (the word "disaster" is from his point of view, I mean it's far from being great, but it's enough for some basic use cases).

Anyway, with more and more urgency coming to get things sorted in Windows world, I'm pretty sure more resources will be attached to fix some Xen parts on that. As he said, an incremental approach will be the solution, testing some patches for Xen devs will be a tremendous help.

That's a very good example on why cooperation is a powerful way to move forward. Thanks a lot for your efforts @XCP-ng-JustGreat 👍

@danp said in XCP-ng 7.5.0 final is here:

@dblogic said in XCP-ng 7.5.0 final is here:

Just tried updating our pool and after updating the master had failed migration of VM (Ubuntu) and then was unable to get VM to start (Missing VDI error). Only solution was to roll master back to 7.4 and everything working again.

I ran into a similar situation when I upgraded to 7.5 this week. Did you check to see if there was a non-existent ISO mounted in the CD rom for the VM in question?

Thank you. That was my problem. I gave cd empty. Everything works.

Actually, looks like I had already written something about it in the Updates Howto: https://github.com/xcp-ng/xcp/wiki/Updates-Howto#a-special-word-about-xcp-ng-75-76-and-live-migrations

anyone can request a change in https://github.com/xcp-ng/xenadmin/issues 🙂

@stormi dd stands for disk dump and does exactly that: Copy a stream of data.
Fio however can be configured for precise workloads and read/write mixes, parallel workloads etc.

So the first thing will only give you streamline benchmarks, what almost nobody cares about.
The second can simulate realworld (VM/database...) workloads, where (controller) Caches and non magnetic storage (Flash, Optane, MRAM...) makes the real difference.
Also use big amount of data, since caches can impact small ones extremely. Don't get me wrong: We need them and they can make huge differences, but as long as your benchmarks fully fit into them, it gives your nonsense/fake results. Also (consumer) SSDs start throttling after some 10 to a very few 100 GB of data written. Their caches fill up and they 'overheat'.

You can spend days on benchmarks and how to do what. 😉

👏 👍 👍

Not my area of expertise. Perhaps this will point you in the right direction -- http://kb.eclipseinc.com/kb/can-i-safely-ignore-io-errors-on-dm-devices/