VMware
 

Knowledge Base

Search the Knowledge Base:
Products:
Search In:
 

On ESX Server Systems with Multiple PERC 4 Controllers, LUNs Become Temporarily Inaccessible After Upgrading the PERC 4 Firmware

Details

VMFS on local storage connected to PERC 4 (4/DC, 4e/DC, 4/Di or 4e/Di) controllers may become temporarily unavailable under the following conditions:
  • ESX Server 2.x with two or more PERC4 controllers (This condition requires any two controllers using the same PCI technology.)
    • PERC 4/Di and PERC 4/DC are PCI-X controllers
    • PERC 4e/Di and PERC 4e/DC are PCI-e controllers
  • Local disks attached to each of the PERC4 controllers and configured with VMFS volumes
This condition can occur when:
  • Firmware for the controllers is upgraded to version 352B (for 4/DC or 4e/DC) or 522A (for 4/Di or 4e/Di) from a previous version.
  • Or, there is an ESX Server 2.x installation with PERC4 controllers at firmware version 352B/522A.

How to tell if your ESX 2.x server is affected by this issue

One or more of the local VMFS volumes becomes unavailable, and output of the commands for local storage is similar to the following examples.
 
Here, vmhba1 is the first PERC4 controller:
 
# cat /proc/vmware/scsi/vmhba1/0\:0
 
Vendor: MegaRAID? Model: LD 0 RAID5 1023G Rev: 352B
Type: Direct-Access ANSI SCSI revision: 03
Id: 3f 39 39 39 39 39 39 39 39 39 4c 44 20 30 20 52
 
In the next command, vmhba2 is the second local PERC4 controller.
 
# cat /proc/vmware/scsi/vmhba2/0\:0
 
Vendor: MegaRAID? Model: LD 1 RAID5 121G Rev: 352B
Type: Direct-Access ANSI SCSI revision: 03
Id: 3f 39 39 39 39 39 39 39 39 39 4c 44 20 31 20 52
 
Notice that the UUID of the local disk is identical for both controllers.
 
The multipath query output would look similar to the following example:

[root@esxhost root]# vmkmultipath -q

Disk and multipath information follows:

Disk vmhba1:0:0 (1,022,998 MB) has 2 paths. Policy is fixed.
vmhba1:0:0 on (active, preferred)
vmhba2:0:0 on
 
Disk vmhba1:1:0 (121,318 MB) has 2 paths. Policy is fixed.
vmhba1:1:0 on (active, preferred)
vmhba2:1:0 on

Solution

The only workaround is to roll back the firmware to version 351X (for PERC 4/DC or 4e/DC) or 521X (for PERC 4/Di or 4e/Di) or upgrade firmware to version 352D/522D (or higher) respectively.
 
Warning: To avoid running into this issue, do not upgrade the firmware version to 352B or 522A.

Also see "VMFS Datastores Are Inaccessible After Upgrading the PERC 4 Firmware on Systems with ESX Server 3.0.x" at http://kb.vmware.com/kb/1001577.

Keywords

Local Storage Problems Firmware Upgrade PERC 4/DC PERC 4e/Di Controllers

Feedback

Rating: 1 - Lowest 2 3 4 5 - Highest (0 Ratings)   

Did this article help you?
This article resolved my issue.
This article did not resolve my issue.
This article helped but additional information was required to resolve my issue.
What can we do to improve this information? (2000 or fewer characters)
Submit
Rating: 1 - Lowest 2 3 4 5 - Highest (0 Ratings)   
Actions
  • KB Article: 5966817
  • Updated: Aug 14, 2009
  • Products:
    VMware ESX
  • Product Versions:
    VMware ESX 2.0.x
    VMware ESX 2.1.x
    VMware ESX 2.5.x