FreeNAS destroyes USB Sticks and SSD's frequently

[joeschmuck]

SSD's dead: 2x Crucial MX500 250GB

One of my first SSDs was a Crucial M4 that had a firmware issue, it would stop writing after a specific number of power on hours. After power cycling it would work for about 1 hour and then stop working again. During that 1 hour was the only opportunity a user had to upgrade the firmware but early adopters of this drive were doomed for a while until the firmware fix came out. My drive hit that magic number a few weeks after the new firmware was released and when it happened to me I was quite irritated, it was my Windows computer that I did everything on. I thought to myself that SSD's sucked! Thankfully I found the firmware fix and resolved my problem.

My point here is that SSD's do stop working through no fault of our own. Also infant mortality is a very real thing. I think the firmware problem likely was fixed for the MX500 lineup but sometimes quality control isn't what it should be.

I don't have the exact logs of those, but we are talking about a failed drive after 3-4 weeks being used as a boot pool only drive on a private NAS.

I'd think this was infant mortality, bad luck. Of course as @jgreco pointed out, you could have a noisy power supply (meaning electrically noisy).

Well, I'd mostly been buying MX500 and WD Blue for hypervisor datastores for the last year or two, but this Black Friday, the 1TB 860 Evo's were going for $100 and that was hard to pass up (pointing to the stack of 18 boxes).

I picked up two of these myself as well from NewEgg, good price and I do have 4 Samsung SSDs (850 Evo and 860 Evo) that have been working well and the Samsung products work well for me.

Maybe I'll just order some brand new Intel Data Center SSD's

I don't think you really need these devices but it sounds like this is for an office/work place system so spending the money and having the strong warranty will likely make you feel better. If you have more SSD failures then look at the power supply.

As for tracking the SMART data, you could use the script smart_report.sh that @spearfoot has provided to everyone which will send you an email when it is run which contains all the SMART data you need to track a problem like this. Just have that email filtered and saved in a folder for later use. I have customized my version of this for my specific needs and it works great.

Good Luck

 

[no_connection]

Why? USB sticks haven't been recommended boot media since 9.3 was released, however many years ago that was, and for exactly that reason.

Why are you saying it should kill USB sticks? And that there should be no way to use USB sticks for boot?

Why is it wrong to want to find a way to use USB for boot and save SATA ports that most systems don't have an abundance of. (without paying 10 time to convert SSD to USB)
You can even have a boot loader that just acts as boot and hand off to a volume attached to a regular pool. Or just have every ZFS disc bootable and stop this nonsense of needing a separate boot drive.

 

[danb35]

Why are you saying it should kill USB sticks?

I'm not saying it should kill USB sticks, but I don't believe it's worth investing any real effort into preserving USB sticks, particularly since they've been on the naughty list for over five years now. If you're short of SATA ports, get a HBA. They're cheap, widely available, and very well supported.

 

[SweetAndLow]

Over the last 5 years my usb boot devices have only had maybe a couple hundred gigs written to them(all from just upgrades). I would figure out what's writing to your boot device so much, something is not correct. Either an error in setup or possibly log spam of some kind.

I keep waiting for my usb devices to die because I read all these posts but it never happens.

 

[ChrisRJ]

While I have not been active on the forum for long, I was an occasional reader for almost a decade and that is how long I have been running FreeNAS off USB sticks. One died a couple of years ago and that is in line with what seems to be the overall consensus. I know that "the plural of anecdote is not data". But I would argue that the likelihood of a hardware issue (PSU or otherwise) is bigger than a flaw in FreeNAS. Otherwise, given how widespread FreeNAS is used, similar reports would have come up more often.

Have you considered changing the PSU or motherboard to rule those out? How "clean" is your power? About 20 years ago I had a customer with a NetWare 3.12 server that would crash daily until a UPS was added to the setup. It turned out later, that a building site nearby would cause glitches in the voltage ...

 

[joeschmuck]

Have you considered changing the PSU or motherboard to rule those out? How "clean" is your power? About 20 years ago I had a customer with a NetWare 3.12 server that would crash daily until a UPS was added to the setup. It turned out later, that a building site nearby would cause glitches in the voltage ...

That reminds me of a story in the UK, we had some equipment which continued to randomly fail. We replaced everything, multiple computers, peripherals, everything. Then we started looking at the incoming power. This was a military installation with "clean" power. Well it wasn't that clean and the voltage fluctuated over time. We have three power analyzers connected to various locations and recorded data for about 6 months. The problem were the main transformers for the building, all needed to be replaced. Even a "passive" component can cause problems.