Jump to content

Recommended Posts

Posted

so hyper-v is in, make a new VM from the network... spend 30 minutes trying to figure out why it freezes on PXE... needed to make a new virtual switch.. nice.

 

anyhow I've made a gen 2 VM based on my pupil image, took a while but ok it's running. Start using for a bit ok. Now come back to it an hour later and it's not running anymore? (pc was locked, and went to standby)

 

Now the VM says I have to restore or continue?? ok continue... VM is dead invalid data in boot sausage.

 

How is it this fragile?

Posted (edited)

If the PC went to standby, then that's your problem. Hyper-V is quite robust.

 

If the host PC goes offline (shuts down, hibernates, whatever) then your VMs will pause. The host has to be operational.

 

Edit: Unless I've misread your configuration. Your configuration might help.

 

Most of us run hyper-v for our entire infrastructure and it operates 24/7 without issue.

 

Edit-Edit: A hypervisor can support S3 and S4 states depending on OS etc, but I noticed that you said "PC" and not "Server" so jumped to a conclusion there.

Edited by Mako
Posted
Have you installed the hyper V integration services which would have suspended or shut down the VM before the host slept - assuming that you had set that as an option. Making a VSwitch is just part of the setup process, without being rude it really sounds like you might need to read the documentation a bit more before starting. I have had nothing but good experiences with Hyper V.
Posted

Have to say I find Hyper-V fairly robust, too. Most of my schools have a number of server VMs and a couple of "PC" VMs running 24/7 without those sort of problems.

 

Don't do our images that way any more, but IIRC there was something if you were using the VM after sysprep.

Posted

How is it this fragile?

 

Used it for years.. and rarely had problems.

 

Even when I first looked at it, got it setup and working (self taught before I did some training) it was easy.

 

I originally used VMWare before I went to Hyper V. Both worked and both had no issues.

Posted
yes the host is windows 10 and it was the host that went to standby. I'm just using the default options, I'm surprised that it was able to bork a VM so easily.
Posted

Well that was likely down to the host going into sleep mode that caused the issue, more so than Hyper-V itself.

 

Servers would never be configured to have a sleep setting so that would not normally be an issue, and all of our workstations that we have Hyper-V setup on for testing with VM's also have sleep disabled.

Posted
Well that was likely down to the host going into sleep mode that caused the issue, more so than Hyper-V itself.

 

Servers would never be configured to have a sleep setting so that would not normally be an issue, and all of our workstations that we have Hyper-V setup on for testing with VM's also have sleep disabled.

 

If you have the right integration services installed and the right options in settings, then the host going to sleep would pause or shut down the VM so it doesn't bork on restart.

Posted
yes the host is windows 10 and it was the host that went to standby. I'm just using the default options, I'm surprised that it was able to bork a VM so easily.

 

I occasionally use it on W10 as host at home, but I've never had the host decide to snooze. I'm surprised it messed up the VM so badly.

Posted

The fact you said about "Now the VM says I have to restore or continue??" sounds more like you haven't disabled the automatic checkpoints, as I'm assuming that restore comment meant revert or continue

 

In Windows 10 (rather than server) HyperV has automatic checkpoints turned on by default (at least in the newer updates of Win10), if you interrupted the checkpoint merge/deletion when it went to standby it could easily upset secure boot because that's the purpose of secure boot :p And I'm guessing the boot error was then related to that potentially. Personally would disable both parts on Win10 (automatic checkpoint and secureboot) by default when testing if using Win10 etc

 

Steve

Posted
MS have created (yes, bought and developed) a product that (sceptical) IT staff are quick to defend. That shows you how good it is. It’s probably the best MS product we have on site.
  • Thanks 2
Posted
I occasionally use it on W10 as host at home, but I've never had the host decide to snooze. I'm surprised it messed up the VM so badly.

 

Occasionally a power cut or other unplanned shutdown will really do a number on the computer, I hit the powercutton an a 4 way once on my test network and it totally hosed the AD before and since I have never seen anything worse than needing to run scan disk.

Posted
uraghhhh! I've built a VM working fine I've installed WJEC as it's being a pain again. (I like pain apparently) The VM needed a reboot (after .net 3.5) annnnd its black.. HyperV can't shut it down... I've reset the VM and it's still black... I'd disabled sleep on the host so fml? another VM dead.
Posted

UPDATE: it's the HDD!

 

I installed a 2tb HDD in the PC for the VMs use.... it was an accidental purchase for the CCTV, that I then removed, forgot why and put in my desktop.....

 

I removed it as it was a SMT style drive... The drive is 100% usage and is killing the performance of the VM. (I'm only running one VM) but that's enough to saturate the drive apparently. I briefly saw the VM do something whilst stuck on it's black screen.

 

This drive isn't even a year old and apart a brief stint in the CCTV machine (2 weeks) it wasn't used.

  • 2 weeks later...
Posted

In all my years of using hyper-v I have never had a corrupted VM. Never. Ive even made mistakes in the early days (I once restored a checkpointed DC, you learn about rollbacks the hard way). We have a machine on our network at the moment that started life as a physical 2003 poweredge 2800 - yes, I will let that sink in - it was P2V at 2008 then it has been in place upgraded all the way to 2019 at the moment!

 

We have golden samples in VMs gen1 and gen2, we have linux machines, DB servers, web servers, student "web" server for A level ICT that are sandboxed, development test VMs. Infrastructure is a stretch VM cluster.

 

Ive made mistakes on VMs and needed to restore (Veeam is awesome - the only failing being a slow restore), even windows backup and restore is good enough for a P2V! Production checkpoints are great for lab testing - I have even checkpointed an SQL server as a production checkpoint to see if it worked and it does flush the logs properly!

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now



×
×
  • Create New...