Recently, i’ve been planning to build all NVMe based NAS to render the files. Don’t need much storage (under 10TB) but needs to be fast enough. While i was discussing with my friend who is into homelabbing, he mentioned that his one of the Intel D7 P5520 died all of sudden. The SSD was no longer detected in any machine and the BIOS would hang.
I’m aware that there are several form factors when we talk about SSD (M.2, SATA, U.2/U.3, SAS, E1.S, etc) and i’m also aware that electronics can fail but despite these issues, seeing an enterprise drive failure which are build for long term storage, read/write extensive with petabytes of endurance, it is quite disappointing to see these enterprise grade drives fail.
Now, i’m in confusion whether i should go with Intel D7 series or choose a M.2 SSD for this build. I’ve seen like two SSDs (M.2) failure (KIOXIA) which were OEM drives one failed during hardware install and the other failed randomly and i’ve had like more than 100 M.2 NVMe SSDs in office in NAS and main machine, some at home and never had a failure. These are all equipped with WD drives at the moment but i have used Samsung, GIGABYTE in the past.
Any recommendations and any insights from someone who has faced the SSD failures? Especially U.2 from Intel
Infant mortality is a real thing. Corporations buy cases of drives and destroy them when they fail, no RMA. This happens all the time. It doesn’t matter who the manufacturer is, drives fail.
Sorry, I only have limited M.2 NVMe failure experience, which is not saying much.
Yeah, i agree. Yes, new to U.2 but really had very less failure of M.2 (not even close to 5) which makes me think what to go for. Any U.2 failure experience @etorix, @NugentS ?
We just recently had a pair of 15.36TB Micron 9300 U.2 NVME drives fail out of a 5+ year Dell R740xd based TrueNAS core appliance. TrueNAS marked them as predictive fail and one of them began to experience very large numbers of IO errors. We’ve not had a chance to test them.
We also had four 15.36TB Micron 9400 U.2 NVME drives fail out of a pair of PowerEdge R7515 based TrueNAS core appliances. These electrically failed. We’re not entirely sure what happened here as we have a considerable number of Micron 9300 and 9400 series drives deployed for various uses. The appliances that our 9400s failed out of also transit the country via Fedex fairly regularly so it’s possible these are actually mechanically induced. Because they are dead, we don’t really know.
We have approximately 250 x 15.36 and 30.72TB Micron 9300 and 9400 NVME drives deployed in various uses. We’ve been using these models for about 5 years and these are the first failures we’ve seen.
I think 5yrs and having a failure is within the expectations as the warranty is 5yrs on these drives, provided that the machines were in 24x7 operation.
Experienced the same with D7-P5520 my friend had.
Could it be heat? Less likely in your case though.
