@flomer said:

OK, I guess I could try and test it if I have the time soon 😉

Did you get a chance to test?

@scottalanmiller Yeah... I know..

It was just never bothered with. .. 😞

I'm making a snapshot right now using NAUBackup of my VM's before running the upgrades. As ya never know if something will get broken.

Understanding the difference between a good risk decision, and getting lucky....

Youtube Video

@johnhooks said:

This was resolved. The VM was running. After shutting it off, the interface could be removed.

Oh, just saw this. Expected behaviour when running. Same on VMware too, for example.

Exactly - two 6 TB HDD's may be a lot less expensive, but the performance compared to SSD would be insanely different.

How many HDDs would you need to reach your IOPS requirements?

I've never monitored them.

The work is being done by Dom0, not the VM.

And with the resources being statically assigned I can't imagine that there is much of a hit to the performance of the VM's them self.

Also, remember it is XenServer and not Xen server, those mean two different things. I'm assuming you mean Citrix XenServer here?

NFSv4 is only for security, typically you want to avoid this in a production virtualization environment. You would want to be on NFSv3 for performance reasons.

this is the screenshot of my XP machine

@thwr Correct 🙂

@dafyre said:

My thinking was more along @scottalanmiller's this time... But like everything "it depends" on what the server does. For a File Server, or an application server (Exchange, et al) that we previously mentioned, to me an agent would make more sense.

However, I thought that restoring to different hardware was actually a function of the restore process, and not necessarily part of the backup process. (I guess it depends on what backup software we are talking about).

Your first, and thus main, backup should always be the entire VM for anything.

If you then need some of the features a backup from inside the OS gives you, then make secondary backups there.

Veeam can be made application aware and issue commands inside the guest to ensure the specified application is in a state for a valid recovery. I would assume that Unitrends has similar functionality, but no idea.

@art_of_shred said:

There is a very high possibility that the IP it was operating on was the iDRAC's IP. I wonder if the iDRAC console somehow became the active connection.

That's an interesting thought. Not sure how that could have happened but since the IPs overlapped.... hmmm...

@coliver said:

So would this be something like OpenStack or CloudStack? Or are those just a management/obfuscation layer?

No, MaaS is designed to help OpenStack specifically (but you can use anywhere. OpenStack is the cloud layer. This is a hardware management layer that makes tons of sense in a cloud scenario. It could be used without cloud, like for a computer cluster (Hadoop maybe) but would be uncommon. And you normally use cloud without MaaS. But it works really well for clouds to make them less effort to manage at the hardware level.

Not very exciting updates, but I suppose that that makes sense in the hypevisor these days. What exactly would we be looking for?

Quite a while means several months, not years.

http://www.xenserver.org/