Highlighted
jrothwell88
1 Nickel

NS20 critical issue

Background:

I'm a computer systems engineer and a network administrator. I was given this NAS by a company when they upgraded. I am familiar with other NAS servers but I have never worked with EMC. I installed it in my rack at home and got everything up and running.

Issue:

I started configuring the device and long story short, I (stupidly) changed the IP addresses of the SPs via Navisphere. They rebooted and I obviously lost connection to them. After that happened and the NAS Control Station was rebooted, I could no longer access the web management page. I was able to access Navisphere by connecting a laptop to the 'wrench' port on the data mover and assigning the laptop an IP on the same subnet as the new SP IP. I changed the IPs to be on the same subnet as the Control Station via http://[ip]/setup. After that, I could no longer connect to the SPs via Nivisphere with the old or new IPs.

Then I (really stupidly) tried to reinstall the Control Station OS thinking that would allow me to get the web management page working again. Now, I'm stuck with the installation failing because it cannot access the back end storage.

I'm pretty sure now I have to run a back end cleanup on the SPs (and find a way to change the IP addresses back) to get the reinstall to work. I read a post about it but then realised that my partial installation of the control station does not have any of the NAS directories (ie nas_raid or nasmcd) because the install would not complete without the back end storage.

At this point, I cannot access Navisphere or the control station management page. I can get to the serial CLI of the control station and I can ping the control station IP address.

There must be a way that I can fix this, it's just software and configuration settings, I just need some guidance on the process. If I can not find the resources to do it myself, what support options do I have as this device is about 10 years old? It doesn't help that most of the posts I see just end with "call EMC support". Everything I have is still functioning and my SPs are fine so it should just be a matter of configuration.

Thanks to anyone who can help me with this.

Jason

0 Kudos
9 Replies
dynamox
6 Gallium

Re: NS20 critical issue

Jason,

unfortunately EMC never made back-end tools available to customers, so things like re-imaging always required EMC support involvement.  As yo probably guess EMC will not touch this array so your best bet it to contact a 3rd party reseller that might have some in-house expertise.

0 Kudos
PSetiawan
1 Copper

Re: NS20 critical issue

I don't get it... the new ip you change it to doesn't work ?

Have you tried this ?
Reconfiguring Storage Processor (SP) IP address using Navisphere Manager

0 Kudos
jrothwell88
1 Nickel

Re: NS20 critical issue

Thanks. I have reached out to a few third party support companies...any that you would recommend?

0 Kudos
jrothwell88
1 Nickel

Re: NS20 critical issue

Thanks for the reply. Once I changed the IP address of the SP (using the method that you liked to) originally, I could access it via the new IP, that's how I changed it back to what it was before. Once I changed it and the CS was rebooted, I could no longer access the SP by either the new or old IP. I was connected directly to the SPA port on the back-end data mover. At this point, after I made the really dumb decision to reinstall the CS OS, it seems like there is close to no IP traffic between the CS, the client data mover, and the back-end data mover. The link lights flash once every once in a while and the amber ! point is flashing on each one. I connected to the SPA port and ran wireshark to see if I could find what IP it was on and there is literally no traffic coming from it. I know it's my own fault, but this sucks..........ugh.....I'm used to working with equipment where as long as the physical hardware is not broken, the software can always be reinstalled and reconfigured to make it working again.

0 Kudos
Rainer_EMC
5 Rhenium

Re: NS20 critical issue

well - the OS re-install was unnecessary and what got you into trouble.

yes you may need a backend cleanup before a reinstall - there should be a knowledgebase article or Solve procedure for that.

0 Kudos
jrothwell88
1 Nickel

Re: NS20 critical issue

Thank you. I agree, stupid move on my part. I took a look at the backend cleanup but I keep getting stuck on "System 128.221.252.200 not up, keep trying..."

Some parts referenced commands on the CS from the /nasmcd/sbin or /nas directory but with the current state of my CS OS install, those directories are missing. I only have the following directories.

bin   etc    initrd      media  opt   sbin     sys       usr

boot  flash  lib         misc   proc  selinux  tftpboot  var

dev   home   lost+found  mnt    root  srv      tmp

0 Kudos
jrothwell88
1 Nickel

Re: NS20 critical issue

Both data movers are in a reset state. How should I resolve this? I feel like that’s my next step. Thanks again.

0 Kudos
jrothwell88
1 Nickel

Re: NS20 critical issue

I have heard back from a 3rd party support provider, they will be sending me a quote. I just want to make sure that no one has any other steps or suggestions before a drop any money on this....I am fully capable of resolving the issue, I have just never dealt with this type of device before and I'm unable to find the steps/documentation.

I need to:

-resolve the data mover 'reset' state issue.

-re-establish communication with the backend.

-cleanup the back end (I have documentation on this)

-re-run the CS OS install (I have documentation on this)

Again, thanks to anyone who can help.

Jason

0 Kudos
jrothwell88
1 Nickel

Re: NS20 critical issue

****I cannot proceed with the Celerra Control Station OS install because the backend is not cleaned up. How to I clean up the backend without the NAS software installed on the CS?****

Please help.

0 Kudos