LinuxQuestions.org

LinuxQuestions.org (/questions/)
-   Linux - Newbie (http://www.linuxquestions.org/questions/linux-newbie-8/)
-   -   fsck in ext3? (http://www.linuxquestions.org/questions/linux-newbie-8/fsck-in-ext3-16510/)

recurrente 03-18-2002 07:42 AM

fsck in ext3?
 
I have SUSE 7.3 installed, with the root partition in ext3 filesystem. However, when I reboot the machine uncleanly, fsck checks the partition as a ext2-fs (2 or 3 minutes), is this normal?? I have in /etc/fstab that partition as ext3-fs

Thanks.

acid_kewpie 03-18-2002 08:43 AM

i get that too, but never bothered to try and see why... i've had a look just now tho. what does the output from "dumpe2fs /dev/hda(...)" say on the filesystem features and state lines? i get:

Filesystem features: has_journal filetype needs_recovery sparse_super
Filesystem state: clean

for a recovering partition, but

Filesystem features: has_journal filetype sparse_super
Filesystem state: not clean

on my root, which like you, needs to recover long style. I've tried fscking it in single user mode, but it doesn't seem to do anything useful. i'd suggest searching around for those flags and relevant sounding words in google like i have been, without a huge amount of success.

d3funct 03-18-2002 11:35 AM

If you look at the man page for mke2fs in RedHat 7.2 you will see that there is are flags (-j an d -J) for making ext3 filesystems. This tells me that the underlying process for ext3 filesystems is still ext2 fs + a journaling system to make it ext3. So, since ext3 appears to continue using legacy ext2 as a base it makes sense that e2fsck would run to check an ext3 filesystem, but e2fsck would note the filesystem was built with the -j or -J options and apply whatever was necessary to check it as an ext3. Does that make sense to anyone else?

acid_kewpie 03-21-2002 05:15 AM

Quote:

Originally posted by d3funct
If you look at the man page for mke2fs in RedHat 7.2 you will see that there is are flags (-j an d -J) for making ext3 filesystems. This tells me that the underlying process for ext3 filesystems is still ext2 fs + a journaling system to make it ext3. So, since ext3 appears to continue using legacy ext2 as a base it makes sense that e2fsck would run to check an ext3 filesystem, but e2fsck would note the filesystem was built with the -j or -J options and apply whatever was necessary to check it as an ext3. Does that make sense to anyone else?
erm, yeah makes sense... not sure what relevance it has to the question, but it makes sense :)

actually i realised... ext3 is compiled as a module normally isn't it? consequently the / partition can not be checked as ext3 as the module is on that partition, and not in the kernel image... so to get it to go as ext3 you'd need ext3 compiled into the image. ok now THAT makes sense doesn't it? is there a simpler way to do it without recompiling a kernel?


All times are GMT -5. The time now is 12:30 AM.