0.9.9 Causes Disk Optimize to Fail Topic is solved

[Engineer]

With PrimoCache v0.9.9, when I run the "built in" Windows 8.1 "Defragment and Optimize Drives", all seems fine - for awhile. Then a system crash occurs - causing drive corruption. I did create a drive backup before testing - all is good.
With PrimoCache v0.9.8, "Defragment and Optimize Drives" runs through completely without issue.

I did repeat twice on one system and once on another identical system - same results.
Unfortunately the windows Event Viewer shows no errors or warnings (actually a minor miracle in itself).

System:
Widows 8.1 (All latest updates and drivers installed)
Gigabyte GA-F2A85X-UP4 Motherboard (AMD Trinity APU)
RAID 1 (2x Western Digital WD2002FAE drives), 4K Cluster Size, Evenly partitioned as drive C: and D:
SSD Corsair Neutron GTX 240GB SSD

PrimoCache:
Only cache drive C:
Level-1 Cache - 4096MB
Level-2 Cache - 228821MB
Block Size - 16KB
Strategy - Read-data & Write-data
Defer-Write - 10s

I also notice that about one in three shutdowns hang - I must hit the power button to actually get the system to power down. This goes away when PrimoCache v0.9.9 is uninstalled.

[Bjameson]

I can confirm the shutdown problem. Windows 8.1 x32 running off a cached volume often fails to completely shut down. Using a 8192 MB RAM cache, RW, delayed write 300 secs. I waited about 30 min. but Windows will not shut down. I think the cache was flushed because there is no corruption on restart. It appears that Windows goes down up to the point where it turns off the power. It hangs there more often than not. Windows 8.1 x64 always shuts down normally.

I have the same problem with XP running on a Virtualbox VM. The host is cached, the guest is not. It always hangs at the same point, just before power off. Restart instead of Shutdown always works normally. Same cache setup as described above. The cache remains running on the host and shows no problems.

Using an Asus P6T-SE mobo, 36 GB of RAM. i7 CPU, Samsung 840 SSD, Vertex1, Vertex2, WD 2TB spinner.
I tried to run the VM both from the spinner and from the Vertex2. On the spinner (also cached) it shuts down normally. On the SSD it always hangs.

In my case, it's enough to disable the cache. I do not need to uninstall Primo 0.9.9. Maybe it's a Windows problem made worse by the cache. Because I have another machine with XP and no cache. This fails to shutdown maybe 1 in 50 times. Maybe Windows is too fast now, or something. It's not a problem for me but it may be fatal in a production environment.

[Bjameson]

Update: I managed to solve my shutdown issues by uninstalling Primo 0.9.9, re-installing it and re-installing Windows. It's completely stable now. No hung shutdowns anymore.

It makes me wonder how many "bugs" are in fact caused by frantically installing/uninstalling/re-installing various software versions without cleaning up first. I am guilty of this but next time I'll take more time and do my testing in a more structured manner.

My sincere apologies to Romex. Primo was not to blame. It was my stupid fault.