Yes. Summarizing the info, no drive had an increase in faults from one round of tests to the next. Some did have a number of bad sectors that were flagged in the first test and re allocated but they did not gain any additional from subsequent stress tests. I also did test with 2 different systems to try and rule out blind spots in the testing methods.did you check the smart data
I was going to, figuring that I could pull logs when I rebooted and got back in but TrueNAS is, well, I have never been so disappointed in software before now. TrueNAS seem to move the system data CONTAINING THE LOGS off the boot drive(which iXSystem strongly recommends using a high quality SSD so it can handle the constant writes of configs and THE F-ING LOGS) and on to the data pool. So I have nothing to fall back to as it crashed on startup and given that there is nothing on this yet, wiping and starting over is easier than trying to recover log files.post the actual error
I have pulled the 2 pools with the greens and dumped some test data on the pool made of purples and it looks at this point that lack of (OR presence of, these greens were in NVRs before my possession so they may have had it manually enabled) TLER was/is the issue.(?, I am still not sure. Will test further with a few other drives that are practically brand new and post stress test but lacking TLER to see if that was it. )
According to additional digging on the subject ZFS auto marks a drive taking too long to respond as having a bad sector. Is there a config option to adjust the timing on this in TrueNAS? I can't see giving the drive no time to self correct as a good thing when the drive manufactures are building better and better self repair in the drives. Unless of course speed is the desired result, but then who would be using spinning rust for that anyway?