FreeNAS ASRock C2550D4I Experience

[FreeNASBob]

I know there have been a lot of reports of this issue lately, but I wanted to add my experience here so that others can benefit from my mistakes.

I built a FreeNAS box using the ASRock Rack C2550D4I board back in 2014. Just like so many others, my board lasted about 2 years at which point it would not boot any more. I thought it was the watchdog bug, but it turned out that the voltage sensors had failed (another common thing with these boards). Luckily I live close enough to ASRock that I went and exchanged my board in-person. As luck would have it, the board they gave me wouldn't even POST. At that point I was thinking, "Awww hell, it's something in my setup and not the ASRock." I was wrong. I got out the PSU testing equipment and verified the PSU checked out. Tried different RAM. I even tried no RAM just to see if I'd get an error. Nothing. The board simply won't even power on at all. IPMI works, though, and you can see the same symptom as my original failed board: some of the voltage sensors report "Not Available".

So not only did ASRock suffer from a major product defect, but they're replacing these RMA boards with more defective boards. There's no way they even tested this board before giving it to me. They would have seen right away that it was also defective. I'm starting to wonder if somebody at ASRock shorted Asrock, Inc. stock and is now solidifying their position. I don't know why else you'd willingly sabotage any good name you had in the industry by refusing to address MAJOR bugs in your hardware and even replacing defective hardware with more product you already know is defective.

I think the sum of people's experiences with ASRock at this point justifies steering new users away from ASRock hardware for mission-critical applications. If there was ever a need to make sure that a company appeared on some "Not Recommended" list of hardware for FreeNAS, this is it.

 

[m0nkey_]

The C2550D4I boards have been rather unfortunate, however their C22x and C23x boards have been the opposite and perform very well. Maybe adding ASRockRack to the 'not recommended' list is a touch harsh, maybe just the Atom C2x50 boards.

 

[Arwen]

Many computer manufacturers have had lemon products. Here is a very partial list from my memory;

• Seagate and IBM both had hard drives that had glue like lubricant, preventing spin up, (before 2000)
  
• Seagate's 3TB SATA drives have a very poor reliability reputation, (probably deserved)
  
• Intel's famous FPU bug from the 1980s
• Sun Microsystems Ultra SPARC II 400Mhz ECache parity errors, (sun spots, bad manufacturing, ...)
• Recently Samsungs phone problems, (catching fire, melting, ...)
• Most early SSDs from ALL manufacturers had poor endurance, failing long before they should have
• Intel's Itanium CPU. While HP used it, as did a few others, no one else adopted it

Please note that I have a Asrock Rack C2750D4I board in my FreeNAS Mini. It's still working fine. As soon
as the watchdog bug was reported, and work around given, I implemented it. Hopefully it will last the 4 more
years I was desiring.

 

[Ericloewe]

Intel's famous FPU bug from the 1980s

90s, it was the Pentium

Recently Samsungs phone problems, (catching fire, melting, ...)

Don't forget their exploding washing machines.

 

[FreeNASBob]

I don't want to come across as unreasonable. Of course unforseen things happen. Bugs creep in. Parts are unexpectedly defective. I get it. It's the response that matters. In this case ASRock is dishing out more bad parts and so far unwilling or unable to produce corrected parts. We don't know because they aren't saying.

I have far more respect for a company that says, "oops, that product is messed up. Here's what went wrong and what we're trying to do to fix it," than I do for a company that tries to sweep it under the rug.

So far I've been without a FreeNAS system for 17 days, 16 of which are the result of waiting for ASRock. It would have been far longer if I didn't live close enough to exchange the product in person.

FreeNAS needs to design something like RAIDZ for motherboards. :)

 

[m0nkey_]

I don't want to come across as unreasonable. Of course unforseen things happen. Bugs creep in. Parts are unexpectedly defective. I get it. It's the response that matters. In this case ASRock is dishing out more bad parts and so far unwilling or unable to produce corrected parts. We don't know because they aren't saying.

I have far more respect for a company that says, "oops, that product is messed up. Here's what went wrong and what we're trying to do to fix it," than I do for a company that tries to sweep it under the rug.

I just think you had a bad experience with ASRockRack and are looking for somewhere to get it off your chest. Once the bug was identified, a workaround was offered. I do believe that iX Systems reached out to ASRockRack when the problem was discovered as this board is used in the FreeNAS Mini pre-built systems.

I on the other hand have not experienced any issues with ASRock.

 

[FreeNASBob]

I just think you had a bad experience with ASRockRack and are looking for somewhere to get it off your chest. Once the bug was identified, a workaround was offered.

There is no workaround for a bricked motherboard and a DOA replacement. Everyone has their requirements and tolerance for risk, so for some people it's no big deal to get the features and take on the risk of having to go through a few weeks of RMA process if they get unlucky. Other people might be more risk averse and would rather choose different hardware. I am just reporting actual facts of what happened with ASRock hardware and how the RMA process went. It's a single data point for anyone considering ASRock hardware for their FreeNAS build.

If I wanted to get things off my chest it would be a little more of a rant.

I on the other hand have not experienced any issues with ASRock.

And that's great. There are a lot of people who haven't and there are a lot of people who have. If it's worth it to you to save some money and take on the risk of becoming one of the unlucky, that's perfectly fine. It's not an invalid choice, but it's an informed choice. I'm just doing my part to enable people to make an informed choice so they're not surprised if they lose the gamble.