topic Re: 3par 7450 issues in HPE 3PAR StoreServ Storage

3par 7450 issues

efirkey — Fri, 20 Apr 2018 15:27:54 GMT

Hi Everyone.

I am new to 3par and I just received a used 7450 4node 3par system and I have no support. Currently, the system has no drives in it. I connected by console to the system through the MFG port using putty and the special adapter.

The system is cabled correctly but all the interconnect nodes are amber, so I figured there must be a communication problem between the nodes. I then logged into each node (cli when I could using 3paradm 3pardata) and checked out the node identifier and netshow information. This is what I had.

16xxxx5 is the serial number
Node in slot 0 is 16xxxx5-0 ipaddress 130.175.93.202
Node in slot 1 is 16xxxx5-2 ipaddress 192.168.1.206
Node in slot 2 is 16xxxx5-0 ipaddress 192.168.1.206
Node in slot 3 is 16xxxx5-1 ipaddress 192.168.1.206

I moved the node in slot 2 (16xxxx5-0) to slot 0, node in slot 3 (16xxxx5-1) to slot 1, node in slot 1 (16xxxx5-2) to slot 2, and node in slot 0 (16xxxx5-0) to slot 4 (because it had a different IP address then the rest)

When I rebooted the system all the interconnect connections are green except the ones from node 3 ( the node with the different IP address)

There is no cluster. It appears two nodes think they are node 0. And one of the 0 nodes has a different IP address.

Currently, I can log into nodes 0, 1, and 2 with 3paradm 3pardata, but I can not log into node 3 with 3paradm 3pardata even though i could before.

How do I get the node currently in slot 3 16xxxx5-0 ipaddress 130.175.93.202 to have the same IP address as the others and become 16xxxx5-3?

I would try a node rescue but the node in slot 3 thinks it's node 0 and has a different ipaddress. That's why I want to change the ipaddress and make it node 16xxxx5-3

Will a cluster form without drives in the systems?

Pretty much all the console (console cmp43pd) commands hang or do nothing on node 3, so I can't change the IP address there.

Thank you for any help and please let me know if I need to clarify anything.

Re: 3par 7450 issues

Venkatesh_SL — Tue, 24 Apr 2018 12:21:49 GMT

Please refer to the below link:

3par 7450 best practices

https://h20195.www2.hpe.com/v2/GetPDF.aspx/4AA4-4524ENW.pdf

Product details:

https://h20195.www2.hpe.com/v2/GetPDF.aspx/c04111384.pdf

Re: 3par 7450 issues

vibeesh1 — Wed, 25 Apr 2018 04:30:37 GMT

Hello

From the Provided Information,

Node in slot 0 is 16xxxx5-0 ipaddress 130.175.93.202
Node in slot 1 is 16xxxx5-2 ipaddress 192.168.1.206
Node in slot 2 is 16xxxx5-0 ipaddress 192.168.1.206
Node in slot 3 is 16xxxx5-1 ipaddress 192.168.1.206

i suspect that the "Node ID" in the PROM settings and slot positions need to be checked and corrected if it set with wrong Node ID vs Slot ID. In a proper System (Bottom to Top)

Slot 0 -->Node ID 0

Slot1-->Node 1D 1

slot 2-->Node ID 2

Slot 3 -->Node ID 3

The above can be checked in the Whack Prompt and can be corrected accordingly. (Contact HPE for assisatance).

Once the above is confirmed yo need to perform OOTB(out of the Box) to make a customized storage as per your requirement

Register a Kudo by clicking the thumb if this helped in your issue.
Please consider marking it as an Accepted Solution if issue is resolved.

Thanks!
Vibeesh V

Re: 3par 7450 issues

efirkey — Wed, 25 Apr 2018 14:03:07 GMT

In case someone else runs into this problem. I roughly did the below procedure. I take no responsibility if this doesn't work for you or if you lose data.

I did move the 3 similar (ipaddress) nodes to their correct slot id node id positions. I put the odd ip addressed node 0 to slot 3 went into whack (cntrl-W during boot) and changed the node id and ip address. Next I powered up the three bottom nodes, did a deinstallation (setsysmgr wipe 16****5) and restarted the system. I then issued a startnoderescue -node 3 command. I immediately plugged node 3 back in to power up and start the rescue. I plugged the console into node 3 and I could see the rescue while it progressed. Rebooted. Ran setsysmgr wipe again and then all the nodes joined the cluster and all nodes have the correct serial number and node id.