• We’re currently investigating an issue related to the forum theme and styling that is impacting page layout and visual formatting. The problem has been identified, and we are actively working on a resolution. There is no impact to user data or functionality, this is strictly a front-end display issue. We’ll post an update once the fix has been deployed. Thanks for your patience while we get this sorted.

850 Evo slow to initialize? 4x850 Evo on FreeNAS pool

V

Viper GTS

Lifer
Odd issue I'm seeing with a SSD based ZFS pool:

I have a 3 week old pool of 4x850 Evo in a stripe of mirrored vdevs under the latest build of FreeNAS. It's running on an older LSI based onboard SAS controller flashed to IT mode, then a SFF 8087 split out cable. As soon as I yank my 16 1 TB SATA drives they'll get a newer SATA 6Gb controller + SAS backplane but for now they're running at SATA 3G speeds.

The very first time I powered them on i noticed that the detection order was odd - Three SSDs were detected sequentially, then all my spinning drives, then the last 850 Evo. It worked fine, and has ran flawlessly for the last 3 weeks.

Yesterday I restarted FreeNAS to run updates and it came back with one drive missing. I started an rsync to mirror that datastore off onto spinning disk until I could figure out WTF was going on. Thirty minutes or so into the copy I realized that the drive was back and working fine again. FreeNAS brought it right back into the pool and the pool returned to healthy status. Looking at the console of the VM you can see the drive pop in as if it were hot inserted.

I've power cycled the entire machine several times since then and each time the same results - One drive comes in at least a few minutes after boot. It's always the same drive, and once it's there it's fine until the next restart.

Where should I be looking for solutions to this? I'm going to replace the controller and all cables in a day or two, but it's odd to me that I don't see it drop out - It just seems like it takes it anywhere from 5 to 30 mins to start functioning after a restart, then it's completely fine.

Anybody seen anything like this? Any suggestions on what I should be googling that will result in useful data?

Viper GTS
 
I've seen a fair amount of issues with SSD's and old controllers. That said, given it's the same drive every time, it seems it's unlikely that it's the controller. Do you have another break out cable you can try? Are the drives all running the same firmware version? If you pull the drive and toss it in your windows box, does the BIOS detect it right away?
 
Haven't tried a new breakout yet, though I do have some extras on hand.

Working on burning in some new 6 TB disks so when that finishes I can replace it when I have it open for the rest of the new controllers.

Trying to find other people with similar issues but just get a lot of 'my new ssd isn't as fast as I think it should be' garbage.

Viper GTS
 
I've got two SSD's (Crucial MX100's) on my ZFS box, the boot drive and the ZIL. IBM M1015 (Cross flashed to LSI9211 IT) plugged into an LSI expander. No detection issues. What onboard controller do you have?
 
Onboard was a LSI 1068.

I have now migrated to all LSI 2008 series controllers and backplane connectivity, all appears well now. It would come in late every time before, I've now done four boots with no issues.

Maybe it was a bad cable (since three of four worked fine) but it's strange that once detected it was apparently completely stable.

Viper GTS
 
You must log in or register to reply here.

TRENDING THREADS

Back
Top