Dell Vrtx And Quality Control

[PliotronX]

So one of my company's largest clients had brand new Vrtx servers deployed not two months ago and two blade modules crapped out and a daughter board in the CMS gave up the ghost. Anyone else seeing this kind of problem with them? $30k servers just dying randomly? Then there's little issues like the HDD activity lights depending on tiny mirrors to route to the front of the chassis. My coworker spent a couple hours trying to figure out why one disk wasn't illuminated like its brother in RAID10... not impressed Dell 🙁

 

[drebo]

Hmm. Not encouraging. We just bought one of these for a customer.

 

[seepy83]

My take - every manufacturer is going to run into QC problems at some point. Within reason (i.e. very small % of product with problems), the difference between a company I want to do business with and one I don't is how they respond to the problem. What did Dell have to say about it?

 

[PliotronX]

Hmm. Not encouraging. We just bought one of these for a customer.

You can probably have confidence in the unit, just keep in mind that when things go wrong they can really go wrong but...

My take - every manufacturer is going to run into QC problems at some point. Within reason (i.e. very small % of product with problems), the difference between a company I want to do business with and one I don't is how they respond to the problem. What did Dell have to say about it?

True, of course they are going to stand behind it and overnighted new modules but still it crippled a major part of the network and in fact took out an office site entirely for a day and half. The clustering setup did keep most of the company going but we would have been screwed if the other Vrtx went AWOL 😀

I guess I just find it humorous and ironic that all of the health monitoring hardware involved is useless when it all just craps out, in this case caused by the very hardware meant to foresee failure... Dell violated the golden rule of KISS.

Oh and as for what Dell responded with, I was not involved in the replacement process only in the initial config as I couldn't travel out of town because I would have missed too many class sessions (finals was last week). I only answered support calls from the office with the outage and escalated it to our server specialist who determined the failure cause.

 

[drebo]

My biggest concern with the unit is the fact that there isn't a redundant backplane/controller for the storage subsystem.

That said, they are an interesting solution in the SMB space.

 

[Emulex]

all the mighty have fallen due to roached firmware (dual controller/dual backplane).