SSDs are dying one a time - trying to find the culprit

[drobo]

They have a 540TBW value according to this document. Which is also the reason I was interested in understanging the OP's use case.

Btw, @drobo you did not answer the question regarding the trim cronjob: not trimming the drives might cause issues.

Apologies I missed your edit about usage, and totally missed the question on trim.

As for Trim, I have Auto-trim disabled on my pool and I currently don't have any jobs that run besides some daily backups and scrub/SMART tests, is that as simple as enabling it under the pool settings?

 

[Davvo]

As for Trim, I have Auto-trim disabled on my pool and I currently don't have any jobs that run besides some daily backups and scrub/SMART tests, is that as simple as enabling it under the pool settings?

Yup, just enable auto-trim. Some users prefer to have a daily/weekly zpool trim poolname cronjob due to an issue quite a few versions ago IIRC.

 

[drobo]

Just an update, I moved my NAS to my workbench to start working on it, I had to power it up so that I could check that it was still reachable via my local net and I'm seeing all my disks back... this is the first time the formally sdq drive has been back online in quite a while. I am now seeing 2 drives with Checksum errors now which I know was mentioned earlier:

I'm going to start a full backup tonight to my old synology that I got back online and hopefully be able to start looking into the possible power issues tomorrow.

EDIT: As suspected the same 2 disks have faulted overnight with read & write errors, and now a third disk has started read errors :( I'm really hoping this backup finishes before anything else faults

 

[drobo]

So my current PSU is a Corsair CX750M, and actually seems to have a pretty beefy 5V rail (from what I understand SSD's only use 5V?), it can push 25A on the 5V, which is actually higher than the current 850 and even 1000w Corsair PSU's which (if the sources on the internet are right), only push 20A max on the 5V. So I'm a little lost on what to do next.

I suppose I will bite the bullet and buy a new 1000w PSU, at the very least it will let me use dedicated cables for each drive since Corsair cables are supposedly all standard and both these PSUs both use "type 4" corsair cables, so I should be able use my old SATA cables with the new ones on a single 1000w PSU to at least eliminate the Splitters..

Definitely regretting going budget with lots of smaller SSDs, should've just gotten 4TB or 8TB drives :/

Edit: Actually guess I'm getting an 850W, the 1000W is too long to fit in my case :D

 

[Davvo]

The real issue is that I can't seem to find the specs regarding your SSDs' power draw.

 

[drobo]

The real issue is that I can't seem to find the specs regarding your SSDs' power draw.

Same here, I've emailed back and forth with Teamgroup's support a couple times due to RMA's so I've sent them a message to see if I can get any info.

 

[drobo]

Not to bump my own thread again, but I finally just finished replacing the PSU, man did I have everything run super tightly haha.

Completely forgot that in addition to the splitters, I had one SATA power extender, which basically added 4 more sata power connectors to the end of a 4 sata power cable... I had to remove some of the cables to see all the labels on my SSD's, but I'm fairly certain all of my currently AND past failed disks were connected to that particular cable in one way or another..

I now have 5 separate non-splitter, non-extended, SATA power cables connecting all of my drives, and on boot... I have no errors. 0 on every drive.

I'm going to let this burn in for a while on my workbench to make sure its good, which I'll report back on, but either way, big thanks to everyone who helped me up til now and gave me things to try and to think on. I've learned more about PSU's and PC power in the last couple days than years of DIY building!

EDIT: I suppose I am going to call this solved, got my nas back in place and still 0 errors on all drives, thanks again all!