Originally posted by DanL
View Post
Announcement
Collapse
No announcement yet.
EXT4 Lets Us Down, There Goes Our R600/700 Mesa Tests
Collapse
X
-
It's becoming more tempting to switch to XFS, fall-back to EXT3, or just eagerly await the stabilization of Btrfs.
Seriously, if you are doing benchmarks with graphic cards you know you are bound to get lock ups once in awhile...
Comment
-
I'v had similar problems. Did a fresh install, upgraded the kernel to 2.6.32, tried some experimental radeon thing that hard locks, and had many files in many directories get screwed(including entire folders), along with an unbootable system. FSCK was going crazy, basically repeatedly fixing thousands of things (probably breaking it more). I don't understand how something like that passes QA testing. I was using a fakeraid setup, do not know if that makes it related or not.
Comment
-
Originally posted by figvam View PostIt's suspicious that the whole directory was purged. Could it be your testing software bug?
Have you checked the lost+found directory? Have you checked the fsck output in the boot log?
Comment
-
Originally posted by Michael View PostUsually it serves its purpose fine and worked well back when the problem with EXT4 was the empty files on crashes.
nice std mount options for /<root> and /home as well
...
But saying that my "backup" actually revolves around critical data mirrored on my desktop as well as every 3months burnt to DVD's
but keeping the data in the same folder, on the same partition, on the same drive, in the same machine sounds a much better backup scheme. I shall be employing such a scheme immediately
Comment
-
How about a Phoronix article comparing backup solutions for Gnome and KDE that are being actively developed?
Like Back-in-Time, TimeVault (active?), sbackup, git :-), etc...?
Not really useful for a "I write my own solutions"-nerd but definitely interesting for the average user.
Comment
-
Originally posted by howlingmadhowie View Postand with a time machine, the probability is either 1 or 0
Heh, on second thought, nope. The probability of data loss is equal to the probability that the resources required to run the time machine are less than the cost of reproducing the data.
Comment
Comment