pid 45645 (vmware-checkvm), uid 0: exited on signal 10 (core dumped)

[mha]

Hi,

I am running FreeNAS-11.1-U7 on a HP Microserver Gen8 (Intel Celeron G1610T, 2,3GHz, 2GB RAM) with 3 2GB disks in a ZFS raid. It is used mainly as a file server plus PLEX in an iocage. Everything running good and stable. I can’t really say when but believe it started after upgrading from U6 to U7 to occasionally throw a message in the security log. I searched but couldn’t find any plausible reference to this message. My installation does not run in a VM so I assume some application is trying to find out if it runs in a VM and crashes. I’d very much appreciate your assistance.

Code:

freenas.local kernel log messages:
pid 45645 (vmware-checkvm), uid 0: exited on signal 10 (core dumped)

-- End of security output --

Thanks,
Michael

 

[dlavigne]

2GB RAM

You need at least 8GB for FreeNAS and at least 16 GB is recommended if you're also running Plex.

 

[mha]

Think this is the root cause for the log entry? My 2GB system us up an running for 4yrs now.

 

[dlavigne]

Yes.

 

[mha]

what does the message mean? what exactly is happening?

 

[jgreco]

You're probably pushing the system way far into overcommit and causing swapping, and there are too many things that want to be in core, or maybe a few really big things that want to run in core. Some of the VMware-supplied stuff isn't the greatest, either.

You should be aware that there is a history of people having significant trouble with ZFS and FreeNAS on systems with less than 8GB RAM. In some of the more severe cases, people lost data or even their whole pool. It's one of those things where "it worked great" for people until one day something happened like they lost power or had a disk fail, or sometimes just came right out of the blue.

It's 2019. RAM is cheap. Please stick a little more in there, 16 at a minimum, but Plex will be happier at 32 especially if you have a large database.

 

[mha]

Ok, will check for more memory. Unfortunately it won’t accept more than 16GB and it needs ECC RAM. Not cheap.

Thank y’all for your support.

 

[mha]

I upgraded to 16GB. System ran for 15 days without problems til today with same message in the security log:

Code:

freenas.local kernel log messages:
pid 44655 (vmware-checkvm), uid 0: exited on signal 10 (core dumped)

-- End of security output --

Memory status:

 

[cJZ]

The checkvm binary comes with open-vm-tools... That is meant to run inside a VMware guest. If your appliance is running on bare metal, why are you using open-vm-tools?

Code:

pkg remove open-vm-tools-nox11

will likely solve your problems.

 

[mha]

I hadn’t chosen to run VMware deliberately. Is it installed as part of FreeNAS 11.1?

The only thing I manually changed was creating one iocage and installing PLEX.

 

[cJZ]

I hadn’t chosen to run VMware deliberately. Is it installed as part of FreeNAS 11.1?

The only thing I manually changed was creating one iocage and installing PLEX.

I guess it does, I just verified my appliance by running pkg info | grep nox11 and it showed up. I'll leave this one to someone more experienced.

Another suggestion would be to upgrade to 11.2-U4 now that you meet the RAM requirements.

 

[mha]

Thank you very much for your support!

 

[mha]

Upgrade to 11.2-U4.1 succeeded. I‘ll wait and see ...

 

[mha]

A few days after I updated to 11.2-U4.1 U5 became available. So I updated and am happy to report my system is up and running 11.2-U5 for 31 days now without the vmware-checkvm error message returning. To recapitule: I ran 11.1-U7 with 2GB and saw recurring vmware-checkvm error messsages in the kernel log messages. After advice here I upgraded to 16GB memory which didn't solve the problem. Updating to 11.2-U5 ran free of any errors for 31 days now. So I guess my problem was more related to a 11.1 bug.

Thanks to Y'all for providing great support.

 

[jgreco]

The checkvm binary comes with open-vm-tools... That is meant to run inside a VMware guest. If your appliance is running on bare metal, why are you using open-vm-tools?

Code:

pkg remove open-vm-tools-nox11

will likely solve your problems.

This is FreeNAS, an appliance, not FreeBSD, a general purpose operating system. You're not supposed to be adding and removing things to the base system.

Open-VM-Tools is included because this is an essential component for virtualized FreeNAS instances to be able to shut down by direction of the hypervisor. Open-VM-Tools is a bit flaky though.

Do not try to remove it.

 

[mha]

Not removed

 

[cJZ]

This is FreeNAS, an appliance, not FreeBSD, a general purpose operating system. You're not supposed to be adding and removing things to the base system.

Yeah, that's what this was about. I may need a bit of RTFM.

"I guess it does, I just verified my appliance by running pkg info | grep nox11 and it showed up. I'll leave this one to someone more experienced."

Open-VM-Tools is included because this is an essential component for virtualized FreeNAS instances to be able to shut down by direction of the hypervisor. Open-VM-Tools is a bit flaky though.

Do not try to remove it.

Thanks for chiming in about open-vm-tools jgreco.

I then proceeded to suggest an upgrade, which seemed to have fixed OP's issue.

I'm also aware that suggesting an upgrade is also not likely the best course of action, as now we don't know the root cause, but either way.

 

[jgreco]

Oh, we know the root cause... open-vm-tools is flaky. I've been fighting that particular issue for years.