Reply With Quote???Originally Posted by jumico
In my opinion the Pentium M test is uninportant and SSD hardware is too rare among normal users, so far.
But I could be wrong.
Senior Member
It's not. It's comparing ext4 with btrfs on an SSD, and comparing ext4 with btrfs on an HDD.
Junior Member
???
In my opinion the Pentium M test is uninportant and SSD hardware is too rare among normal users, so far.
But I could be wrong.
Senior Member
Huh? The point was to compare the filesystems on a fast machine and a slow one. Mission accomplished. The idea is not to specifically compare to the hardware you actually have, but to show the behavior of the filesystem in such a way that you can extrapolate the results to anything. If a filesystem led these tests on both machines, it's a good bet it will always do so no matter what the hardware is. If it differed, then you can look at the results and try to split the difference depending on what kind of hardware you have and which system it's closer to.
It might be nice to throw a medium one in there as well, but this is a million times better than the older reviews where Michael would pick just one.
Last edited by smitty3268; 06-27-2011 at 01:20 AM.
Junior Member
I think it would be more meaningful to benchmark btrfs when it's actually completed. Is there any summary of what has yet to be done for btrfs anywhere? As soon as it's done I'd love to switch from ext4 to btrfs.
Senior Member
I have both, a pentium-m laptop and a laptop with SandyBridge i7 and SSD
The only problem I have with Linux 3.0 on SandyBridge is a dma lockup, if I have a constant stream of data (like copying >=3GB over the network). This is since 2.6.39.2, input devices do not get any proz time until the copy is finished or interrupted (hitting ctrl+C permanentally works after ~1min).
