Esxi 4.1 eventual adaptec raid hang

[MarcusFPereira]

I have a system with VMWare Esxi 4.1 (and all patches). The hardware is:

. Dual Intel Xeon-Westmere 5670-Hexcore [2.93GHz] vmware.

. 48Gb RAM

. Adaptec 51645 Raid Controller

. 12 Western Digital 250GB Hard Disks in several RAID 10 arrays.

I am having occasional datastore hangs, like 1 each month.

During the hang I can access the host on vsphere client or ssh but I cant access any virtual machine. Vsphere fails to do any operation on the virtual machines. Listing directories /vmfs/volumes hangs at the ssh session.

Today I have around 20 virtual machines at this host but may be only 5 have heavier disk usage (nothing too heavy). But I had this issue even when this server had only 10 light virtual machines running.

Anyone with similar problem and/or solution?

[AndreTheGiant]

Welcome to the community.

I do not use your RAID controller, but I suggest to check for BIOS and firmware update.

And also, if your system is in HCL, consider to use VMware support.

PS: thread has been moved to ESXi forum.

Andre

[MarcusFPereira]

I've updated the Raid Controler to last firmware version (build 18252)  at feb 5.

Just had another crash.

Still no solution.

Next step I will apply 4.1U1 patch.

[MarcusFPereira]

Even with the update the server keeps hanging at random intervals, 1 week to 1 day.

Disabling the Adaptec controler write cache seems to have solved the problem. The server is up for arround 3 months with no new issues.