To actually shutdown a Cx array you need to physically turn off the power switches on both Standby Power Supplies. This sends the array a message to begin the shutdown process and then powers down the Storage Processor Enclosures and the Vault/OS Disk Array Enclosure. There is not a GUI or CLI option for this that I am aware of.

In the EMC Navisphere Command Line Interface (CLI) - release 29 manual - page 289:

shutdown

Shuts down and powers off the storage system

Description The naviseccli shutdown command shuts down and powers off the
storage system you specify.

Syntax
shutdown is used with naviseccli (described on page 41) as follows:

shutdown [-o]

Prerequisites

For naviseccli, you must have a user account on the storage system
on which you want to execute the command.

Options -o

Executes the command without prompting for confirmation.

Output None if the command succeeds; status or error information if it fails.

Example

naviseccli -h ss1_spa shutdown
This command shuts down and powers off the specified storage
system.

Before you shutdown the storage system, stop all I/O from
any servers connected to the storage system and save all
unsaved data to disk. During the shutdown process you
will lose access to all data on the LUNs but you will not
lose any data. When you power up the storage system you
will regain access to the LUNs.

glen

ps - someone needs to read the CLI manual tonight from cover to cover

D'oh! Didn't know that changed in Release 29!

I read the Navisphere CLI Guide for Release 26 (which is what a Cx3-80 would be running) and it states that the 'shutdown' command is only supported on AX series arrays. So is this now supported on Cx series or only Cx4 arrays running FLARE 29?

Looks like CX4 and AX 4-5 only - R29 only runs on CX4 - it was a surprise to me also. There are also two other commands - shutdownsp and shutdownpeersp

glen

Looks like I read the right CLI guide at bedtime after all

OK, so my question is "Why the heck would you want to shut down a CLARiiON remotely anyway?"
To me, unless you are physically there (in which case you can just let the array shut itself down following the normal process) you aren't going to be able to bring it up again. I can't really think of a good case for using this command.

I was wondering the same thing as Allen. Those are some good examples, thanks for your thoughts.

Always happy to help! As a final warning, I don't recommend you put your SPS on a remote powerstrip: an attendee at the 2007 USENIX conference told a story about a disgruntled employee who powered off a CX300 200 miles away and changed the admin password on the remote strip! They lost several days of work sorting the mess out.

I can think of many examples. Recently, a contractor caused a leak in the cooling system for the colo data center. We are all 40mins or more from the datacenter and we had 90 degree temps on the floor by the time the data center manager confirmed the status. I shutdown all of the servers - and even my Celerra - in less than 15 minutes, but I had to have someone enter the room to flip the off switch on the CLARiiON. She was not happy to have to enter a 100+ degree room.

Those of you who are involved with DR and pandemic planning also know that your data centers can be rendered off-limits (no physical entry), if the site is quarantined for viral outbreaks. That means nobody can get in to throw a switch. The same is true for those of us who cleaned up after Katrina - if the roads leading to the data center are flooding or flooded out, the National Guard and state troopers aren't going to let you on the road.

Those are valid point which I hadn't considered. I guess it's one of those things that you never think of until you have to deal with it yourself.

Thanks for the feedback. I always like hearing about real world examples of how things unexpected can affect you in ways you never considered. It helps keep the mind open