[SOLVED] 9.1.0.1.306 vmware hanged system

[daffy]

I'm about to set up NAS storage on a VMware vSphere 5 host and thought I should give Nas4free a chance, but unfortunately the system seems to hang within a couple of days.

Information about the installation:
Version 9.1.0.1 - Sandstorm (revision 306)
Platform x64-full on Intel(R) Xeon(R) CPU L5410 @ 2.33GHz

Running on a virtual machine in VMware vSphere 5 configured as:
Guest OS: FreeBSD (64-bit)
CPU: 4vCPU (2 virtual cpus x 2 cores)
Memory: 6144MB
VMCI: enabled
Scsi Controller: LSI Logic SAS
Network: E1000
Harddrive 1: 1Gb
Harddrive 2: 800Gb

I configured the zfs storage and created a couple of datasets, created a user and enabled the FTP service.
Did some initial testing, uploaded / downloaded ~700Gb data and was pleased with the results (much faster than Freenas8). Returned a couple of days later to do some more tests but didn't get any response from the GUI or from FTP, so I went to the console to check for error messages but the console was completely frozen.

A little confused I killed the power and rebooted and it seemed to work just fine. Searched the logs but didn't find anything. Today I returned to do some further testing, but found out that the server was "dead" again.

I wonder if you have any ideas about what could cause my Nas4free hangs, or if you have any advice about how to debug the problem?

-edit-
(I don't have any problem with other virtual machines (windows/linux) on the vmware host)

Best regards,

[daoyama]

Hi,

Did you try LSI logic parallel?
If you did not use ZFS kernel tune, try this:
ZFS kernel tune (WebGUI extension) 20120920 or later.

[biggsy]

Might be related to a problem reported in the pfSense forums:

http://forum.pfsense.org/index.php/topic,54206.0.html

[daoyama]

CPU: 4vCPU (2 virtual cpus x 2 cores)

http://forum.pfsense.org/index.php/topic,54206.0.html

Thank you. It seems HPET on ESX is buggy until pached above.
Yes, I'm using 1 core CPU on VM. It uses TSC for timer.
If you used 2 core or more, it uses HPET by default.

Try following from shell:

# sysctl kern.timecounter.hardware=ACPI-fast

[daffy]

Thanks for you replies and suggestions. Last weekend I updated the ESXi host to 5.1.0 (build 799733) and haven't had any freezes after that.