thespooler

Hmm, well it doesn't appear to have worked. Preclear is hung at 98% on the first of three cycles' preread stage using the built in preclear script. How should I recover from that? The "/tmp/preclear_stat_sd?" file has a time stamp of 3 hours ago (the preclear itself reports 6 hours have passed, but I started with a reboot and my system uptime is 10 hours. Top shows preclear with 99% usage though the uptime is only 187:33. Is that the missing 3 hours? Perhaps the preread did complete and the next phase is where it hung. Pre-Read (1 of 3): 98% @ 126 MB/s (6:30:25)

I'm trying to do a preclear in a VirtualBox VM on Windows. I have a trial unRAID, with no config or key, unRAID has started. I have installed the Preclear plugin. The HD is passed through and recognized with VirtualBox's RAW mode. Preclear is running, but there is no SMART info available. Am I just wasting my time with this? Or will Preclear still be able to tell me if the drive is okay through actual drive errors it might encounter? I can always check the SMART stats once it's hooked up natively or through Windows. The drive is a NewEgg HGST refurb (supposedly EOL from a data center) and it passed all the SMART tests with HGST's WinDFT.

I had a preclear going (3 passes), it was 3 days in, and I had a UPS event and everything shut down. When everything came back up, there was no record that I can see to just resume the preclear. Are there logs somewhere? I do see "/boot/preclear_reports", but it only has my initial preclears from months ago. If I'm starting over again, is 3 passes excessive, having likely made it to the 3rd pass already? Or would 1 pass be pointless, since maybe it wouldn't know of any reassigned sectors from the intital pre clear.

So how do you know when the files listing memory is getting freed up? Is there a log entry indicating the memory has been released? Or is there a log entry that the file system has been read, which would give you an indication that they were flushed?

My system is only using user shares which I understand are already managed in memory. Does that make Cache Dirs unnecessary for those types of set ups?