Error during initialization of VNX5500

Are you using a version of the VIA that matches the code on the VNX system that you are trying to initialize?

todd

you don't have any virtualizaiton products installed on this system where you are running VIA? (VMware, VirtualBox)

It is a new setup - no hosts at the moment. File OE version & VIA version: 7.0.40-1

Thanks for the tip Thomas. Yes, we too have already followed this document- so we have the assuarance that we are on the correct path.But for some unfathomable reason, we get the same fault after resetting too - "unable to detect the VNX model". We also tried to set the ethernet settings through express install itself - then after detecting the NAS boot LUNs it comes on & says that LUN 0 needs to be zeroed out - our current challenge is how to "zero" LUN 0. Unisphere doesn't allow us to disengage boot LUNs from the "~filserver" storage group, even through engineering mode. 

i- Is there any mechanism of zeroing LUN 0 or disengaging it from SG ?

ii- Has anyone encountered this error while resetting ?

As a follow-up: When you got the “model” error with VIA, did you make absolutely certain that the cabling was all correct? If so, and if you do go down the path to do a fresh reinstall, I might suggest a later version of code then to get away from the 7.0.40.1.

Also, you might check the startup_wiz.log on the VIA client, sometimes it gives a clue as to what might be triggering the model error

todd

To perform the reinstall, you need to be able to run the nas_raid –s cleanup script from the Control Station, using the primus steps in emc257758. If you are already trying the nas_raid and it cannot cleanup and zero out LUNs 0 & 1, then it gets much more difficult.

You will need to perform a new install, and at the point where it finds that there is already data in LUN 0, you say “No” to exit the install. Then, hopefully you can perform the nas_raid cleanup from that point. Unfortunately, there is no other way to zero out control luns, no other tool, etc., and you cannot remove these luns. The only other way to zero out the luns is the more extreme reimaging of the array, which by default will then zero out the luns.

Hope that helps,

todd

The procedure prescribed in emc257758 was followed. We took the precaution of continuing with 7.0.40-1, since opting for a newer version could present us problems with incompatility with the Block OE, which 5.31.000.5.509 at present.

The reset went succesfully, with a single error "The storage array  CKMxxxxxxxx requires security credetials". This didn't appear to be a show stopper, as we rebooted the CS & it came to stage of prompting for the VIA.See attached description of error.

But on VIA we still get the same error. As for Todd's concern about cabling. It is a factory cabled delivery. We tested through pinging to SPA & SPB which respond. Should the Blades respond at this point ? They don't respond to any of the IPs in /etc/hosts of NAS.

startup_wiz.log: does the file get erased after we close VIA, for we couldn't locate it.

I am really running out of options, and the customer is getting upset !!!!

Repition of the steps worked. Thanks everyone.