VM's Lock Up

[Tango2018]

Hello,

Can anyone help if possible, please?

I have a HP Z620 with 3 internal SSD, running the latest Windows 10 and 196GB of RAM and a 6GB Nvidia graphics card in used as a host.

I have created 3 virtual machines which have Windows 10 (latest version), 28GB of RAM with each VM hosted on a separate drive.

The machines are used by my development team using Eclipse to write simple code in, however, we are having an issue where our VM's continually lock up for a minute or so and are very slow. There is no antivirus or security on the system, I have checked no defragmentation and the system load is very low. If I take the same VM's and put them onto another machine they work OK.

[Logs attached - Particularly see around 25/02/2019 10:30-11:20am logs when 3 VMs were active]

Can anyone help, please?

Thank you for your time and suggestions

Message was edited by: Liam *** These are host logs ( I think), please see reply which has the logs from 3 running VM's ***

[Tango2018]

Update: VM logs themselves attached here, I think last logs were from the Host only - sorry my mistake!

These are the logs from the VMs where they lockup please see 10:30 to 11:30 where this happened many time. They locked up, were slow, and showed a black screen and did not do anything for a minute or so

[Susie201110141]

Thanks for posting your issue here, and we will take an investigation of your logs, then update you later!

[continuum]

> however, we are having an issue where our VM's continually lock up for a minute or so and are very slow.
I am not surprised at all. You assigned so much RAM to your VMs that they have to be slow as most of the vRAM will be allocated inside the vmem-file.

OvhdMem_PowerOn: initial admission: paged   943076 nonpaged	18401 anonymous	31517
VMMEM: Initial Reservation: 3878MB (MainMem=28672MB)

I would expect that you get radically better performance if you reduce the vRAM to a size of 4 or 8GB per VM.
If you really think you need 28 GB of vRAM for your VMs you should use ESXi.

[Tango2018]

Hi,

Thank you very much for taking the time to reply.

I have used VMWare for a number of years and to-date never had an issue with the amount of RAM allocated. I have tried reducing the RAM during the testing from between 8GB through to the top amount which was 28GB as mentioned. The workstation the VM is running on has 196GB of RAM. The reason the RAM allocation is high, is the programs developed on the VM require a minimum of 8-16GB of RAM, so taking this into account when pushed to the limit and the minimum requirement for Windows and Eclipse the IDE the allocation appears suitable.

We also have an HP Z600 with a similar setup (inc RAM allocation) and it is not experiencing any of the issues which the Z620 is experiencing. In fact, we can take the VM on the SSD from the Z620 which is problematic, plug it into the Z600 and it works OK.

I have looked into ESXi as you have suggested which appears to be installed like a VM itself and acts more like a 'server' but I can't see what benefits this would provide over a normal VM - could you point me to any useful comparison document, please? Also, when purchasing VMWare told me that a setup such as mine suggested with RAM up to 32GB (and even more) was supported so long as the host machine had enough to allocate and of course RAM in reserve for itself to run the host OS itself. I'm not sure if this has changed - someone from VMWare commented previously but has not since responded?

Really appreciate any help or further thoughts, and I'm happy to set up an ESXi if you feel this would work? Thank you

[Tango2018]

Hello, thank you very much. Just checking to see if there are any updates or suggestions? Thanks

[Eris2323]

I am having a lockup which seems much like the one described in this thread.

I have a brand new HP Z840 running Vmware 15.5.0.  Host is running Windows 10 and the guest is Ubuntu 14.04.

We are seeing system locks ups (Vmware Guest and Host not responding) and we have to hard power cycle.

The VM is allocated 16 cores and 32G RAM (host has a total of 40 Cores and 64G RAM).

The lockups occur randomly.  Usually when large compiles are going on in the Guest.  They also happen when the Guest is mainly idle.

I am not 100% sure we seeing the same issue as described in this message, but also see similar messages like:

2019-10-16T14:16:40.275-05:00| vmx| I125: scsi0:0: Command WRITE(10) took 1.235 seconds (ok)
2019-10-16T14:16:40.275-05:00| vmx| I125: scsi0:0: Command WRITE(10) took 1.234 seconds (ok)
2019-10-16T14:16:40.275-05:00| vmx| I125: scsi0:0: Command WRITE(10) took 1.234 seconds (ok)

Not finding any errors in the vmware.log or WIndows system/application event log right around the lockup.

Was there ever a resolution for the lockup described in this thread?