VMware Cloud Community
StephanGr
Contributor
Contributor

FC problem with ESXi5

Dear holders of knowledge!

I came across a FC problems while trying to get a NetApp FAS 2240 with 8.1.3 7-mode working in a vSphere 5.0 environment.

The storage connects perfectly to the SAN, the zoning is correct and follows the single-initiator guideline. However the ESX hosts were not able to see the newly added targets.

Doing a reboot of the ESX hosts or resetting the fabric port made them see the new target and the LUNs. All paths were there and Alua worked fine then.

Not a big thing until now but then I did some controller failover/giveback tests. The hosts survived the failover without problems but did not see the dead paths as active again after the giveback finished. Again, rebooting or resetting the FC port made the hosts see all paths again.

I found a command echo „scsi-qlalip“ > /proc/scsi/qla23xx/* which triggeres the HBA to login to the fabric again and after that the new targets/revived paths immediately showed up without rebooting/resetting.

Fabric OS is 6.2.2.b (RSCN are enabled)

FC HBA is QLA2312 with firmware 1.54

Hosts are HP BL25p G2 in a p-Class Blade Center

Another host with the same HBA is running Windows Server 2008 R2 and after updating the firmware to 1.54 every things works as it should.

Anybody any ideas what could be behind this issue of not seeing changed targets in the fabric?


Thanx and greetings,

Stephan

Reply
0 Kudos
0 Replies