Re: ext2 question

[Andrea Arcangeli]

On Tue, 18 May 1999, Manfred Spraul wrote:

>AFAIK, NTFS implements journaling with a logging fs:
>they implemented 2 filesystems: the actual NTFS, and a logging fs
>(called LFS) which is stored in a fixed size file on the volume.
>I guess that the LFS is only written (and read during fsck).

I guess the LFS is used _only_ for the metadata.

>The wraparound is a huge problem: you might recall the one of the tweaks
>which Mindcraft did on NT was:
>'set the size of the log to a incredibly large number' (65 MB).
>(the default should be somewhere around 4 MB, but the system gets
>nearly stopped if the log overflows)

For curiosity: do they defrag it while the system is idle? They should
try in all ways to avoid the log to wrap. Otherwise they are forced to
wait for I/O completation of the wrap-fault-handler.

But the log-wrap is not the only problem, the other performance problem of
a real logfs is that we are not going to always read _data_ (not only
metadata) in the same order we write things to disk... And we are going to
read _far_ more frequently than writes...

Anyway a logfs still seems to me something of fun to play with ;).

Andrea Arcangeli