Freenas 12.0 U8
I have two mirrored boot drives, 32gb each. They are ATP CG eUSB cards that plug directly into the (tyan server) mainboard. The sockets they plug in to on the main board are internal and meant expressly for boot devices, they are not just a "regular usb connector meant for anything usb". These are NAND/MLC DOM (disk on module) flash drives.
Since installation I have had the following issue. When I create a mirrored set of the two boot drives, the mirror setup goes fine. Smart tests are all fine. Everything runs fine for a week (exactly) till my pool scrub fires off on sunday night. At that point, the second boot drive goes awry and the pool enters degraded state. If I reboot the server, it finds the second drive in the boot pool and resilvers just fine. It will run fine for a week till next pool scrub and the process repeats. If I never reboot, it just keeps running fine off the 1/2 mirror.
I replaced the second ATP CG eUSB card with a new one.... same problem occurs.
I have a 3rd eUSB card, but haven't tried it yet.
My question is: is it really possible 2 and maybe 3 cards are really bad? If the card was failing, I wouldn't think the resilver would go fine and run well for a week. Should the boot pool drives be left out of the scrub schedule? I'm not sure where to start looking on this, any help would be most appreciated. Thanks!
Best,
Jay
I have two mirrored boot drives, 32gb each. They are ATP CG eUSB cards that plug directly into the (tyan server) mainboard. The sockets they plug in to on the main board are internal and meant expressly for boot devices, they are not just a "regular usb connector meant for anything usb". These are NAND/MLC DOM (disk on module) flash drives.
Since installation I have had the following issue. When I create a mirrored set of the two boot drives, the mirror setup goes fine. Smart tests are all fine. Everything runs fine for a week (exactly) till my pool scrub fires off on sunday night. At that point, the second boot drive goes awry and the pool enters degraded state. If I reboot the server, it finds the second drive in the boot pool and resilvers just fine. It will run fine for a week till next pool scrub and the process repeats. If I never reboot, it just keeps running fine off the 1/2 mirror.
I replaced the second ATP CG eUSB card with a new one.... same problem occurs.
I have a 3rd eUSB card, but haven't tried it yet.
My question is: is it really possible 2 and maybe 3 cards are really bad? If the card was failing, I wouldn't think the resilver would go fine and run well for a week. Should the boot pool drives be left out of the scrub schedule? I'm not sure where to start looking on this, any help would be most appreciated. Thanks!
Best,
Jay