Storageworks 4214R

[James D. Stallard]

Hi

I have a Storageworks 4214r disk shelf. Single PSU, EMU and dual bus Ultra2 controller card, driven by a Smart Array Controller 5302-256Mb.

Recently, all of the disks in BUS A have been suddenly going offline. The drive lights go out and the array (obvously!) dies. The solution is to power off the server and the shelf and restart everthing - whereupon it all comes back to life, until the next time. This usually happens about once a day.

So, several questions...

1. Any ideas why this is happening, I have reseated all the removeable bits, including all the disks. All the LEDs on the back of the unit are always on (EMU, PSU, Blower, controller card)

2. What do the 3 LEDs on the back of the controller card mean (they are all lit all the time).

3. Why don't I get any error messages anywhere, even tho I have all the foundation agents installed, the latest ACU and insight manager v4 running

4. What is the EMU actually doing and where can I get more information on what it does and how to get information out of it.

5. Any suggestions on diagnosing the problem?

Help!

Thanks in advance
Cheers

James

[Leon Rosier · Accepted solution]

Hi James,

I suggest you have scsi connection problem. So the i/o module would be the part to focus on. I would switch around the cables at the back of the module and see whether the fault stays at the same drives or moves over to the other drives in the enclosure.

Your questions:

2. Status LEDs
Left +5.0V Power
Middle SCSI Bus A Termination Power Presence
Right SCSI Bus B Termination Power Presence

4. The Environmental Monitoring Unit (EMU) supports the automatic monitoring of specific enclosure and drive functions. On the front of the chassis are three status LEDs, two green LEDs indicating power status and enclosure status which is normally ON. The amber LED indicates a enclosure fault symbol occurred. A constantly flashing LED providing heartbeat signal is generated by the Environmental Monitoring Unit.


The EMU provides the storage enclosure with the ability to detect and report error conditions about the hard drives, power supply, thermal conditions, and fan operation. The EMU can also monitor its own thermal status.

There are two types of EMU Modules:

Environmental Monitoring Unit (EMU) , Nile Bus Compatible 123481-002
Environmental Monitoring Unit (EMU) , Fault Bus Compatible 123481-003
The -002 module was used on earlier units. (As spares are depleted, substitute with replacement spare 123481â 003).
The -003 module is used on more recent units and is multi-vendor compatible.

5. See above

Le

[James D. Stallard]

Leon

Thanks for the suggestion and the information on the EMU.

I swapped the cables connecting the Array Controller to the IO Module and the problem is unchanged. Suggesting as you said that the IO Module is at fault.

Is it possible that the backplane is damaged? I don't want to buy a new IO Module if the problem is internal to the shelf itself.

Thanks again
Cheers

James

[Leon Rosier]

Yes, this could be the case, but the chance is higher that the problem is in the i/o module. I guess you have no possibilities to "borrow" an I/o module from another shelpf for troubleshooting?
Check E-bay for a cheap replacement..
some examples:
http://cgi.ebay.com/ws/eBayISAPI.dll?ViewItem&category=44990&item=5717574641&rd=1

http://cgi.ebay.com/ws/eBayISAPI.dll?ViewItem&category=44990&item=5717898966&rd=1

http://cgi.ebay.com/ws/eBayISAPI.dll?ViewItem&category=51249&item=5717618244&rd=1&ssPageName=WDVW

Leon

[James D. Stallard]

Top idea, I actually found an U3 IO module for sale, so I'm going to try that.

I'll let you know how it goes.

Thanks once again.
Cheers

James

[James D. Stallard]

ok, brand new IO module in place.

Same problem occurs, all disks in BUS A suddenly power off.

I have swapped disks, cables, IO Module, SCSI card and even swapped the PSU with the blower. The only thing left is the backplane - which I removed this evening and thoroughly checked and cleaned.

Is it possible the unit shuts down the bus if it overheats - the room isn't airconditioned and the rack does get pretty warm, but no higher then about 25 celsius?

In desparation I am considering a new backplane and the rear of the rack has been left open :)

Any thoughts?
Cheers

James

[James D. Stallard]

ok, it's been a week now and the disks baven't died once. It seems that cleaning the thick dust off the backplane and leaving the rear door off the rack have done the trick.

I believe the eventual problem was overheating.

Thanks for your help
Cheers

James

[john m_6]

Dear hp user What you NEED to do is this, each month is to strip down your unit, ie to take out the emu & any removable controllers & psu`s
& clearne of any dust or dirt found!? PLUS if your PSU`s have removable fans such as the EVA 4000 URL ref# http://www.mitlimited.com/html/HP_EVA5000.jpg You will see the psu`s & on URL ref# http://www.itspares.com.au/image/cache/133518-004-120x120.jpg You will see the psu with NO fan? so when you have taken of the fan you need to make shur it is clear of all dust & clean the fans!? so your system keeps NICE & COOL in tip top running order, & as for how you should have it URL ref# http://farm3.static.flickr.com/2668/3985139064_856a538cef.jpg
That info should keep your rig going for years! AND on the mater of the > EMU <
if you get a orange light that is a failure indicator warning?? for help PLease try HP URL ref# http://h20000.www2.hp.com/bizsupport/TechSupport/Document.jsp?lang=en&cc=us&objectID=c00026406 That will give you The added info you need, as in the LED`s ...
I hope this info help`s & if you need more PLease do ask & i will get back to you asap.

[john m_6]

EMU monitoring functions URL ref#
http://h20000.www2.hp.com/bizsupport/TechSupport/Document.jsp?objectID=c00365591#A7