Re: RAID controller failure, or?

I remember having a drive whose serial started appearing with leading 00, i
cant remember the exact situation but it reeks of a bent pin in a SCSI
connector. check the cables that plug into the 2748 for bent pins. For all
the diagnostics codes the i have i lost 4 hours during a cage upgrade just
for that reason (bent a scsi pin when plugging it in the dark...)



On Sun, Jan 26, 2014 at 12:37 AM, Clay B Carley <cbc@xxxxxxxxx> wrote:

Hi DrFranken,

I do indeed have the sidecar, and all of the disks report similar
errors. I'll double check the connections, however nothing has moved
for months; that doesn't mean much though.

The first port on the RAID card is connected to a cable that's zip-tied
to the rear of the chassis, and the middle port on the RAID card has a
cable that's more or less floating inside.

The tape drive is connected to it's own card ( I don't recall the FC
number ) that lives in the sidecar, so it's on a different controller
altogether than any of the disks or the CD drive.

I'll do some more troubleshooting tomorrow then.

Thanks!
Clay Carley

On 01/25/2014 07:23 PM, DrFranken wrote:

So in the system unit there is a cable from the first port on the RAID
card (Rear port of the card) to the drive cage. Then it goes to the CD
and the tape drive. There needs to be a terminator at the end. I can't
remember if the cable is 'all in one' (that is Card to cage to cd to
tape with terminator) or is multiple pieces. Can you verify that all
connectors are on tight and the terminator is in place?

Also recalling that there is a cable that continues on to a hidden scsi
port on the back of the machine verify that it's not been damaged or
gotten wet or corroded. That also would be bad for the connection.

And yes it could be a bad backplane in that cage.

Do you have a sidecar with 6 more drives in it? If so are the drives
there also super slow and full of errors?

- Larry "DrFranken" Bolhuis

www.frankeni.com
www.iDevCloud.com
www.iInTheCloud.com

On 1/25/2014 9:48 PM, Clay B Carley wrote:

New battery arrived today, and I installed it in the 2748 card. Card
went back in the system, and well things aren't looking that good.

Something is definitely wrong with the entire DASD subsystem; all of the
individual disks show 00 as the first two of their serial, although none
of their serials actually start with 00. Attempts at initializing
individual disks results in an I/O error. I also attempted to examine
the disk surface, and the read errors add up incredibly quickly before
any percentage complete occurs.

I even attempted to reclaim IOP cache storage; with the same results.

I suppose *all* of the disks could be bad; but that seems unlikely. I
think I've read about the disk cage backplane going out; I'm sure that
could cause some strange issues.

Breaking down the issue, it's got to be one of a few things, right?
RAID controller, cabling, backplane, or disks?

The tape drive still works great, but that's on it's own controller as
it's an external unit.

I guess it's time to find some more replacement parts.

Clay


On 01/19/2014 12:34 PM, Clay Carley wrote:

I pulled the card out, and found a Sanyo cordless phone battery in

there

with a second label saying it was an IBM part, with the correct part
number. So, I found a battery on ebay (new) for $25 and free shipping
and ordered it.

I was attempting to read the hex dump, but haven't gotten too far with
that yet.

http://carleynet.net/familypictures/view_photo.php?set_albumName=album70&id=IMG_20140119_090724_260

Viewing the full size pic (it's quite large) makes it easy to see

what's

listed. I'm hoping that something will decode to "failed cache

battery".

Now, I do have the old card from the system still, a 2741. Is it worth
throwing that back in until I get the battery for the 2748?

Thanks,
Clay

On 01/19/2014 06:11 AM, Clay Carley wrote:

Hi Jack,

Thanks for the reply. I was hoping that these symptoms pointed to a

bad

cache battery; but seeing 27489010 I was kind of freaked out that
something more had failed.

I purchased the card on ebay last February, and the condition was
"seller refurbished"; which could mean that they pulled it from a

scrap

heap, cleaned it up, and sold as is. So, who knows how old that

battery

really is.

I guess the next step is to pull the side off and see about a new

battery.

Thanks,
Clay

On 01/19/2014 03:50 AM, Jack Kingsley wrote:

I am guessing you have a bad cache battery. I had a similar problem

a long

time ago, got a new system, did some work on it, then let it sit for
several months. How old were the batteries that were in it.


On Sat, Jan 18, 2014 at 7:44 PM, Clay Carley <cbc@xxxxxxxxx> wrote:

To add to this, upon trying to install the LIC to the last remaining
disk, I get:
B6005120

http://pic.dhe.ibm.com/infocenter/powersys/v3r1m5/index.jsp?topic=%2Fp7eai%2FB6005120.htm

Clay

On 01/18/2014 04:34 PM, Clay Carley wrote:

Greetings,

Sadly I've had my 9406-170 off for a few months, as I haven't had

a lot

of spare time to work with it.

Today I figured I'd finally IPL it and start working with it again.
Unfortunately there seems to be a failure inside the box, and I've

spent

the afternoon trying to track it down.

The system is dog slow. I'm attempting to IPL from tape (again),

and

it's still sitting there. Earlier I tried restoring the LIC,

however it

complained that the 2 disks in the mirror set were missing, and I

wasn't

seeing the proper number of disks after pulling those two. At this
point I have 1 disk in the box, and that didn't seem to help.

First IPL failed with code 27489010, which after looking up,

doesn't

sound good. "2748 xxxx It may indicate a system bus

failure."

Will a dead cache battery cause this incredibly dreadful

performance?

Or, is it possible that my RAID card is suffering some failure

that the

system hasn't quite recognized yet?

Thankfully, I have good backups in any case.

Thanks for any tips,
Clay Carley

--
This is the Midrange Systems Technical Discussion (MIDRANGE-L)

mailing list

To post a message email: MIDRANGE-L@xxxxxxxxxxxx
To subscribe, unsubscribe, or change list options,
visit: http://lists.midrange.com/mailman/listinfo/midrange-l
or email: MIDRANGE-L-request@xxxxxxxxxxxx
Before posting, please take a moment to review the archives
at http://archive.midrange.com/midrange-l.

--
This is the Midrange Systems Technical Discussion (MIDRANGE-L) mailing list
To post a message email: MIDRANGE-L@xxxxxxxxxxxx
To subscribe, unsubscribe, or change list options,
visit: http://lists.midrange.com/mailman/listinfo/midrange-l
or email: MIDRANGE-L-request@xxxxxxxxxxxx
Before posting, please take a moment to review the archives
at http://archive.midrange.com/midrange-l.