Command View 7, EVAs 6100 & 4000 : No Replication!

[maxwell_smart]

Hello,

This situation looks sooo strange. The scenario has an SMA using Command View 7.00.01, one EVA 6100 (previously a 5000) and one EVA 4000 both with SW version CR0ECAxc3p-6110. Command View is in its trial period for 180 days, so full functionality should be working.

Both EVAs are almost fully manageable by the SMA: By clicking in the "Data Replication" folder the 6100 displays:
***There are no remote systems visible to the local system.***

By clicking the same folder in the 4000 section, the "Remote System Connection Status" windows only displays the old WWN when this 6100 was a 5000!

What's your feeling about this? Compatibility matrix should be thoroughly explored?

[Vyacheslav Ryzhevsky]

Hi,
what about zoning in your SAN? You need to put your systems in the same zone with each other

[Steven Clementi]

Yep, check your zoning first.

You should have a CA zone.. or 4 zones if you want to get granular...

CA Zone: all ports from both EVA's

or

CA Zone FP1A: the 2 FP1A ports
CA Zone FP1B: the 2 FP1B ports
CA Zone FP2A: the 2 FP2A ports
CA Zone FP2B: the 2 FP2B ports


Your EVA's need to be able to communicate with each other to do CA. In this case, the primary EVA is the "initiator" and the secondary is the "target"

By doing the zoning separately for each port, you can better control where/how your replications works.


Other than that right now... the 2 EVA's should be able to talk and do CA with each other.

Steven

[maxwell_smart]

OK, I've been talking with network administrator and it seems zoning is in its place (it used to work before upgrading EVA5000 to a 6100). He's going to take a look at tne MP Routers where probably there is a "hard" reference to the old WWN of the 5000. I'll let you know if this solves the problem. But before that I was wondering if it would be a good idea to put both EVAs in the same SAN switch, and the same zone obviously, just to discard a firmware issue. What do you think?

Note: the reason both EVAs are linked with a couple of routers is that they originally were in different data-centers. For the purpose of the upgrade/replicate process, the small one was brought to the same site but both routers stayed in the architecture to simulate original conditions.

[Steven Clementi]

It could be posible that the zoning is using the WWPN's of the old EVA instead of the new WWPN's. If this is the case, even though the zoning is "in place".... the zoning is not in place..

Can you telnet to the san switches and run a few commands?

I assume that you have brocade switches, if not, let us know what you have...

switchshow
fabricshow
zoneshow
version


Return the output here.


Steven

[maxwell_smart]

Finally, the SAN administrator reconfigured the extended-SAN devices (MP400) and all came back to normal. As far as I understand, the WWN mapping had to be updated with the new value.

I couldn't find a Brocade tag on the switches, but these are model 283056-B21 and A7987A. Also I wasn't able todo telnet because they're out of my jurisdiction.

Anyway, the final answer was an interruption due to misconfiguration that prevented the tunnel to build-up between EVA controllers.

Thanks everybody for your help!