These fits into the "so frustrating, it's funny" territory.

HP have proven to be no end of hassle when it comes to replacement parts (bearing in mind I'm talking some 5 years ago, things might have changed since). Two specific areas proved to be a problem with them.

The first case relates to product coding. HP sets a product code for memory based on type of chip (DDR2, DDR3), memory capacity (1,2,4GB etc.) and ECC/Non-ECC. That means you can end up with memory chips by two different manufacturers under the same product code. You could even end up with two memory chips by the same manufacturer, but from two different product lines, provided those 3 were the same. Motherboards are unfortunately extremely sensitive when it comes to such things, and they'll either point blank fail to POST reporting no memory, or they'd report the extra capacity but you'd lose Dual-Channel capabilities. We used to hold a bet in the office whether or not an RMA replacement for faulty memory, or an upgrade for a server, would work first time. Eventually we stopped bothering given 95% of the time it wouldn't work. Despite valiant attempts by almost everyone in a department of some 20+ employees over a few years we could never manage to explain to HP just why their product coding system was flawed. Towards the end of my time working with that team the local on-site service rep had learnt to turn up with a good selection of memory chips from the stock he was supplied with.


The second relates to a dying CPU fan. With Dell servers when a component has been dying I've been able to phone up, give the express tag and get a replacement part within the SLA time-frame that has matched. HP doesn't seem to keep the same level of information about their servers.

We had a customer, a major retailer in a particular field, who'd had a few problems due to bad software & architecture choices. Typical over-engineered shopping cart system that used a staggering amount of CPU power for fairly straightforward tasks (makes Magento look lightweight). As winter arrived their servers would melt under the load and even out of the winter season it would be a rare week when we didn't have to deal with one problem or another. After a lot of work with them and their software guys we'd got the solution transitioned to a much more efficient and scalable solution (but one that wasn't scaled just-yet, the plan was for 6 months down the road to add additional DB and Web servers as appropriate), and all was looking great. A whole month went past without having to log in and deal with their servers in any way and we were all rather happy. Then on one of our routine datacentre walkabouts an alarm light was noticed and we discovered that a CPU fan was going on their main powerful dual-CPU database server.
We phoned HP, gave them the serial number and asked for a replacement part.

"What's the product code for the CPU Fan?" came the response.
"Sorry?"
"What's the product code for the CPU Fan? I can't order a replacement if I don't know that"
"Don't you have records of such details?"
"No."

A little peeved we arranged for down time with the customer, got the product code from the CPU Fan, told HP and they arranged for a server technician to come out with the part, and we arranged the second down time for the replacement part.

The part didn't fit.

A replacement part was sent again, and again we arranged down time with a slightly frustrated customer.

That part didn't fit.

An engineer was dispatched with several parts that were supposed to match all the possible ones for the product range. More down time arranged with a slightly more peeved customer.

Neither of those parts fit either. He came back within an hour with one of every CPU Fan from their entire set of supplies, and eventually found one that fit.

In all the process took 2-3 weeks, and 5 lots of downtime, from reporting the problem to finally having a replacement part in the server. Part of the delay was down to arranging downtime with the customer rather than HP. IIRC not long after that we were racking and configuring a hot-spare database server for the customer, who incidentally got through the next Christmas rush without even batting an eyelid. :)

Paul
_______________________________________________
Discuss mailing list
[email protected]
https://lists.lopsa.org/cgi-bin/mailman/listinfo/discuss
This list provided by the League of Professional System Administrators
http://lopsa.org/

Reply via email to