HP EVA6100 Down: How can i get a better support ?

[IT HELP DESK]

Dear members

We moved to an HP SAN recently , we have deployed EVA8100 and an EVA6100 at our office.

Last week we have a HDD failure on one of our EVA6100 and we requested for support. By the time HP send the replacement we have seen multiple HDD failures in the EVA array. This caused around 1.5 TB of data loss.

HP engineer asked to try different trouble shooting tips and collected all logs and send for analysis , but even after 5 days HP is not able to root cause why this happened?

HP support is sending more and more hard disks to be replaced instead of trouble shooting it properly. To my surprise the error from Hard disks are random initially it failed on enclosure 2 and after clearing error bits suggested by HP , now more disks are getting faulted.

Even replaced HDDs are going back to faulted condition immediately.

The last option suggested by HP was to re-initialize the array , so the HP service provider here un-initialized the EVA6100 losing all the data on it. Now the array is not getting initialized back!!!

As per HP i am running HP05 which is the latest firmware for the HDDs.

Can anybody from the forum help me in getting a proper escalation path for Middle east region. or share your experience with HP support.

Thanks and regards
ITHELPDESK

[Patrick Terlisten]

Hello,

what type of support did you buy with the array? Do you have any support contract for it or just standard support? If this array is business critical you should by mission critical or proactive support for it. You will get much better support if you pay for it.

Best regards,
Patrick

[IT HELP DESK]

Hi,

We have bought HA110AC HP Support Plus 24 Service.

Thanks

[Víctor Cespón]

The case should be already escalated inside HP technical support. You don't have a problem with the disks, you have a problem on the loops (cabling / I/O / loop-switches). This is causing communication problems with several disks and the EVA ends up marking them as failed.

There should be someone from second level looking at the logs already.

You can talk to your sales representative and create a bussiness scalation if you're not getting an solution (re-initialize is not a solution, I agree, and is not going to solve hardware problems).

If you have it un-initializaed now, try to use it with only one disk shelf, try each disk shelf separately. This way you insolate the problem. Connect the disk shelf directly to the controller (no loop-switches), like an EVA4100. The less parts on the loop, the better. You only need 8 disks to initialize, if it works add more disks and more shelves. Go up to 4 shelves without the loop-switches. If it works, add the loop swicthes.

It's a lot of work, but after 5 days I don't know why no one has tried this.

[IT HELP DESK]

I really appreciate the time taken to reply on this case.

To my understanding the engineer who claimed L2 gave me the solution of re-initializing the solution and he communicated in mail saying that the data cant be recovered and this is the last solution. So we already un-initialized it.

But still the issue is not resolved , no i am not able to re-initalize it back it fails with disk errors. I even tried initializing with just 8 disks and when it starts initializing then disks are going faulted ( atleast 4) and a red LED is blinking over. Now the commandview says the opration failed.

After few minutes the disks come back to normal.

I already communicated the engineer many times that this does not looks like a HDD issue as the failure comes randomly, it might not be the same HDD which is failing every time!!

Now to my surprise i understood that L2 is still giving all the trials to get it going without trying to get the root cause. the latest solution is to upgrade to CV 8.02 from 8.01!!

I feel we are missing some link in the support chain and we are not getting a proper attention for this case even after 5 days.

I already contacted my sales team and account manager but i am not seeing a satisfying responce, Now actually feeling helpless and regretting for moving to HP platform!!


Thanks and regards

[Víctor Cespón]

That's with only one disk enclosure and 8 disks? Without the loop-switches?

Please understand that there must be some hardware failed, this is not a general problem with EVA6100.

You can initialize it from the WOCP, so no Command View involved, the version is irrelevant. The disks should not go red just like that. Grab the parts you have and try with different disk enclosures and different cables. Do not use the loop-switches.

Last resource is to change the controllers. Also check that the parts are correct, we have had cases of EVAx100 mounted with parts from a EVAx000.

Controller 390856-006
I/O Module A 364549-009
I/O Module B 364548-009
EMU 375393-005

---------------------------------------------

But still the issue is not resolved , no i am not able to re-initalize it back it fails with disk errors. I even tried initializing with just 8 disks and when it starts initializing then disks are going faulted ( atleast 4) and a red LED is blinking over. Now the commandview says the opration failed.

[Uwe Zessin]

> You can initialize it from the WOCP

I've worked with an EVA6100, yesterday, but it did not have a W(eb-based)OCP.

[IT HELP DESK]

Thanks for all your replies,

Now HP L3 support is involved and they are working on this case, I will post the findings as soon as they conclude.

But the kind of support i got in this case is pathetic and the case was with L1 for almost 5 days before they moved to L2 and L3.

And i am checking about the parts as you mentioned regarding EVA-X000/EVA X100

Thanks
IT HELP DESK

[Uwe Zessin]

Right, L1 (L0?, L-1?) sometimes wastes too much time and/or is scared to escalate the case himself.

I beg all my customers to honestly tell about the problems they encountered when asked by HP for feedback about a call. Be polite, but don't just say that every went find just to get rid of the interviewer. Maybe, one day, the person who is responsible for the massive quality drop becomes aware of all the problems and does not just receive the usual 'everything fine' message from the lower hierarchy.
I can dream at least...

[Víctor Cespón]

> I've worked with an EVA6100, yesterday, but it did not have a W(eb-based)OCP.

Too many EVA4400 cases lately ;-)

It's the controller OCP obviuosly.