MSA1000 Critical Lockup Detected Code = f019h

[Sean_Streets]

Hi All,

I was working with one of our healthcare customers here in the UK, and experienced an MSA1000 panic then complete lockup.

I'd upgraded the firmware on this particular MSA1000 from 2.38 to 4.48 a week before.

I had returned to begin zoning up their SAN. I was on-site preparing the aliases on one of the SAN switches, when I noticed that one the MSA's (they have two) had locked up.

I've worked with MSA1000 for quite a long time and never seen this error.

To cut a long story short, tried to get some help from HP, got messed about. Eventually spoke to a guy in Reading, who knew his stuff.

Mentioned that the failure was caused by a Smart Array File Allocation Table corruption.

The fix was downgrade the firmware back to v2.38.

Then make a change to SSP, change the host type from default to windows, then save the config back to the MSA1000.

Reboot the MSA1000, then once up, upgrade the firmware back to 4.48

When I'd rebooted after downgrading. The MSA rebuilt 2 stripes. I then offline upgraded the MSA1000 backup to 4.48.

After the upgrade back to 4.48, as was data was back with the clusters.

General moan really, why do HP Business Partners in the UK get such poor support.

In the old Compaq days, we (ASE's) used to have a seperate telephone number to bypass the 1st line, and talk to the guys who know their onions.

We love the HP Storage Technologies, but I personally hate having to deal with poor support HP have now in place.

All the good Storage ASE know that the only teams worth talking to when this hits the fan are here in the UK, Dublin or Holland.

By the way this whole recovery took about 9 hours..

But don't sacrifice the essence of what Digital, Compaq and now HP is all about.

I.e Quality, Quality, Quality.

Comments anyone?

[Steve Cavendish_1]

Well said Sean. It is not just the UK being handled this way, we have the same problem here in the states. If it wasn't for having very good friends within HP who "accidently" left the numbers of a couple of superstar engineers I would have pulled the few remaining hairs from my head long ago. Unfortunatly the problem seems to be getting wosre instead of better. HP has now outsourced all carepack support levels below 6-hour CTR and to companies who for the most part have never seen a server or storage array before.

When a support engineer comes out to your best customer's site to replace a failed hot swappable IO module on an EVA5000 wearing a security guard uniform and tells the customer they have to shut down 42 servers, the EVA and then unplug it at 2:15am, you know there are problems within the support organization. Even though that might sound a bit far fetched it happened not two months ago. Hopefully it will get better before HP's support organization turns into Dell's.

[Basil Vizgin]

Storage Master ASE's nowdays have direct access to second-line engineers, omitting firs-line support.