Disk Enclosures
1748285 Members
3602 Online
108761 Solutions
New Discussion юеВ

Re: Model 30 "Loop ID: OFFLINE, (1)

 
SOLVED
Go to solution
Jon Mattatall
Esteemed Contributor

Model 30 "Loop ID: OFFLINE, (1)

Silly question time....

We just had a K370/11.00 with attached Model 30 Nike array shipped in for rebuild and deployment. The array should be configured already, so I decided to check.

Everything looks fine...sort of.

The LUNS are as desired, the ENA/RDY looks good, there is no indication of a subsystem fault, BBU failure, etc., and both SP's show the Peer SP as PRESENT.

SP B, however, instead of the expected "LOOP ID: 1", displays "LOOP ID: OFFLINE, (1)".

Huh?

I've rebooted the SP's, reseated the offender, rebooted the box, tried "Soft Enable/Disable". I've checked half a dozen other 30's here, and this is the only one like this. The PROM is at 1.73 for ALL the arrays, so even if it's a little dated, it's consistent across the arrays.

I'm probably missing the forest for the trees, but is this a problem, or one of life's little mysteries that can be ignored? I don't really want to lose a day to rebuilding the array from scratch, but if it's going to save me grief down the line, I will....

Suggestions?
A little knowledge is dangerous - none is absolutely terrifying!!!
4 REPLIES 4
Eugeny Brychkov
Honored Contributor
Solution

Re: Model 30 "Loop ID: OFFLINE, (1)

First idea coming is that there's something wrong with FC link to this controller. Try another FC cable. Try another hub port. Try replacing SP and FC interconnect PCA.
BUT there could be software/firmware problem - are you sure that this controller initializes completely? May it be stuck somewhere? Check unsolicited event logs for both SPs
Eugeny
Jon Mattatall
Esteemed Contributor

Re: Model 30 "Loop ID: OFFLINE, (1)

The forest has sprung into sharp relief...

Booted server into single user mode, changed pw from the previous site's pw, init 3, ioscan.

Nothing on 8/8. No disks. Coincidentally, that's where SP B is connected.

Pull the fiber cable, and use the hi-tech light bulb and eye continuity test. Seems fine.

Grab replacement cable anyway. As I connect it to the server, I notice that Rx and Tx are printed upside down, compared to the other card. But that shouldn't be a problem because the connectors are keyed, right? Right? WRONG.

Sure enough, everything's fine. Some days, It's like playing against a team...
A little knowledge is dangerous - none is absolutely terrifying!!!
Eugeny Brychkov
Honored Contributor

Re: Model 30 "Loop ID: OFFLINE, (1)

Jon,
it's possible to disassemble two-way FC connector and swap wires ;o)
BTW, I did not understand... was problem soled?
Eugeny
Jon Mattatall
Esteemed Contributor

Re: Model 30 "Loop ID: OFFLINE, (1)

Oh, yes. Everything works now.

What happened is that someone (not me, thankfully) switched the cables to put SPA on 10/8 and SPB on 8/8. When they made the switch, SPA was hooked up properly. They then, from what they've told me, hooked up SPB the same way.

The problem was that the fiber ports are on different cards, and on one, the Rx/Tx ports are reversed. The cables LOOK like the A/B lines are keyed, but on the HSC card with the fiber module, you CAN insert them the wrong way. I actually tried this a minute ago to check, and the connector will seat. It takes a bit more force, and it doesn't give out that satisfying click, but it will seat backwards and LOOK fine - right up until you check the SP through the serial port.

A combo of bad luck and being in a hurry.
A little knowledge is dangerous - none is absolutely terrifying!!!