VSAN + DELL PERC: Important Driver and Firmware Updates

I’m currently going through and documenting the build process for our VSAN Management Clusters and one of the first steps I noted down was to double check that the I/O Controllers where compatible as per the VSAN HCL. As I am using the DELL FX2s I checked to ensure that there where no issues with the FD332-PERC (Dual ROC) controller. As shown below there are no issues with it being on the list (confirmed before the actual hardware purchase) however there was a footnote listed next to the Release Info

That link takes you to the MyVMware download for the SAS Driver for the DELL PERC9 Based SAS Adapters, of which there are a number of models listed below.

Version 6.606.12.00-1OEM
Description The ESXi 6.0 driver package includes lsi-mr3 driver version 6.606.12.00-1OEM which enables support for PERC 9 based 12Gbps family of SAS controllers such as Models H730P, H730, H830, H330, FD33xS, FD33xD
Release Date 2015-05-19

I dug a little more into this release and managed to link it back to a VMwareKB (2109665) that talks about adverse symptoms when using PERC9 based controllers and VSAN 5.x or 6.x

  • In the VMware vCenter Server event log display, you see the error:IO was aborted by VMFS via a virt-reset on the device
  • When the VSAN is under load, you can see this or a similar VSAN status display showing VSAN disk(s) unhealthy:
  • High IO latency alarms
  • Failed IO and controller reset messages in the ESXi logs similar to these:WARNING: lsi_mr3: fusionReset:2565: megaraid_sas: Hardware critical error, returning FAILED.
    WARNING: ScsiPath: 7133: Set retry timeout for failed TaskMgmt abort for CmdSN 0x0, status Failure, pathvmhba0:C0:T0:L0

Without copying and pasting the whole KB you want to ensure that you download an install the VIB update listed above and ensure that the Driver and Controller versions are up to date as referenced in the KB. You also need to ensure that the DELL backplane firmware is greater than whats shown below.

  • Expander storage backplane (BP13G+EX): firmware version 3.03
  • Non-expander storage backplane (BP13G+): firmware version 2.23

From the FX2 CMC you can check the versions of that hardware by going to the Update Tab under Chassis Overview, click on one of the servers Update Target in the bottom pane and you are looking for the components highlighted below.

References:

http://kb.vmware.com/selfservice/microsites/search.do?language=en_US&cmd=displayKC&externalId=2109665

https://my.vmware.com/web/vmware/details?downloadGroup=DT-ESXI60-LSI-LSI-MR3-66061200-1OEM&productId=491