EVA6400 LUNs disappear after rebooting BL860c with HP-UX

[Mohanasundaram_1]

Hi Experts,

I have a new setup with BL860c servers connecting to an EVA6400. The Blade has 4GB FC virtual connect to connect to a HP SAN switch.

The LUNs were assigned to 3 * BL860c and after rebooting these servers we found the LUNs disappear. The only way to make the LUNs reappear is to disable zoning on the SAN switch and re-enable.

This works until the next reboot. The same blade enclosure has few Proliant blades with windows which are working without any problem.

We have checked service advisories and we have done a seperate zoning for Virtual connect WWN with EVA ports and another seperate Zoning for server HBAs with EVA.

The FC Virtual Connect firmware is 2.30 and we have HP-UX 11.31 running on all these BL860c servers.

Anybody faced the same problem or have an idea about this problem?

As usual, relevant answers will be rewarded with points.

With regards,
MOhan.

[Mohanasundaram_1]

bouncing this question to the top again.

[Amrutha]

Hi Mohan -

We recieved two Integrity blades and also facing the same problem as you had it. Can I tell me how did you resolve this problem permentantly. If you have your contact phone ,I ca talk to you.

[Mohanasundaram_1]

Hi Amrutha,

Thanks to you that I get to update this thread, which I have forgotten completely.

We solved this problem in 2 steps.

1) We included the Virtual connect switch to the EVA as a seperate zone.

2) We performed the WWN based zoning for all the SAN ports. (Earlier we had used port based zoning for EVA ports and WWN Zoning for servers - By oversight).

I hope you have a similar setup to make this solution work. Please share with us the setup you have so that we can help.

I am in Dubai and so calling me may not be suitable in a mutually convenient time due to time zone differences. But please feel free to post your message here so any body can help you.

All the best to resolve your problem.

With regards,
Mohan.

[attd-storage-l3]

We had the exact same issue with BL870c servers connecting to EVA8400 through 8Gb FC VC with HPUX 11.31.

Our BL460c servers running Windows Server 2003 were able to connect without issue, same as in your situation.

First note that there is Fabric OS compatibility requirements with FC-VC modules. After upgrading our Fabric OS we had better exposure on the switches to the NPIV ports and servers behind them.

We resolved the issue by configuring our WWN zoning with full isolation of each BL870c with the EVA8400. We created an alias for the EVA8400 host ports on the fabric, a separate alias for the BL870c HBA WWN (each WWN configured in separate alias), and separate zone containing the EVA8400 host port alias and BL870c HBA WWN. We have separate zone for each BL870c HBA WWN.

[Mohanasundaram_1]

We solved the problem by doing the below 2 steps.

1) We included the Virtual connect switch port to the EVA ports as a seperate zone.

2) We performed the WWN based zoning for all the SAN ports. (Earlier we had used port based zoning for EVA ports and WWN Zoning for servers - By oversight).

The above two actions resolved the issue.

With regards,
Mohan.