OMG weird, I have exactly the same problem with a agility 2
Tested under openSUSE 11.3 with kernel 2.6.34 and 2.6.35 and the same problem with Xubuntu 10.04 Kernel 2.6.32. Always using ext 4 , installed several times, but sometimes after power off even the whole installation disappeared, I thought I had a defective drive but it seems to be a common problem...
I got curious about the problem...
a little later without power off:Code:dd if=/dev/urandom bs=4096 count=262144 of=/dev/sde ; dd if=/dev/sde bs=4096 count=262144 | md5sum 262144+0 Datensätze ein 262144+0 Datensätze aus 1073741824 Bytes (1,1 GB) kopiert, 329,198 s, 3,3 MB/s 8c29c314237605bb251670ae80e9b524 - 262144+0 Datensätze ein 262144+0 Datensätze aus 1073741824 Bytes (1,1 GB) kopiert, 131,716 s, 8,2 MB/s
But after reboot and power off the md5sum is different, this device is clearly defective and besides brand new:Code:dd if=/dev/sde bs=4096 count=262144 | md5sum 8c29c314237605bb251670ae80e9b524 - 262144+0 Datensätze ein 262144+0 Datensätze aus 1073741824 Bytes (1,1 GB) kopiert, 132,858 s, 8,1 MB/s
Code:dd if=/dev/sde bs=4096 count=26244 | md5sum 262144+0 Datensätze ein 262144+0 Datensätze aus 9d6ee4a89863ec3bea1c7116feea31b0 - 1073741824 Bytes (1,1 GB) kopiert, 134,217 s, 8,0 MB/s
Although I haven't had a change to try a different filesystem, yet this might not be a ext4 issue at all. In this thread: http://www.ocztechnologyforum.com/fo...after-shutdown at least one user mentions he has the same problem on a windows 7 install. Another user has the same problem with the Agility 2. It might be an issue with a bunch of bad Sandforce chips.
Nevertheless I'll try installing ubuntu on my ssd with a different filesystem today.
Update: I installed ubuntu 10.10 beta on an ext3 filesystem on the ssd after I rebooted it was still there, after a poweroff however the ext3 filesystem too was gone...
I wonder how that's even technologically possible. Surely they're not using caches that big in the drive?
What if the problem was that file allocation table (is there such a thing with EXT?) or some other "track 0" data that wasn't properly recorded?
I mostly meant that if data survives a reboot but not a cold boot, this hints into that the data is stored in non-persistent memory.
As I said, it seems likely that this is a manufacturing problem. Apparently OCZ is looking into it right now.