Ron Stodden wrote:
> Civileme wrote:
>
> > Unfortunately, kdiskfree seems to be seriously bugged. I like it and use it in
> > 6.1 but on partitions of 6 G or so I am always shown 1.86G. Another problem is
> > accurately representing nfs mounts .... I am not among those who believe a
> > broken tool is better than no tool at all.
>
> As I informed you when you made this claim before, neither of these
> situations occurs here. I am surprised that once disproved you would
> persist with them?
So if they work for you they work for everyone?
Try
# df /mount_point_larger_than_1.9_G
and see what you get. I did that and got curious results.
Actually, I have 17 machines and kdf on all of them and I am shown a 1.86 G bar for
some single mounts of 1.86 G or larger, nfs or not. I will not persist with
reporting additional problems over nfs because they turn out to be limitations of
knfsd(navigating multiple mount points). But .... HOW USEFUL is it to know that
the 1.86G of a 10 G partition is 54% filled? Now, I am able to find a system where
/dev/hda7 (root directory) shows as 2.84G(correctly) and /dev/hdd1 (/usr/local)
shows as 1.86 though it is 2.08 .... and /dev/hdd5 shows as 1.86G though it is 6.2G
I decided to check the accuracy of the bars, which showed /dev/hdd5 70.9% full
I counted the blocks using a 54 page printout--it was 70.9% of 1.86G
So, could I fill it above 1.86.G? I had hydrogen3 iso files handy and copied
hydrogen3-inst iso into the partition--took it
but 227meg into hydrogen3-ext.iso we had a fatal error--disk full type.
OK NOW I can say kdf is not at fault, but we have one helluva bug in the filesystem
persisting through 6.1, 7.0 and 7.1. This is PARTITIONED for
6388704 blocks and allows use of only about 2000000 of them....
Why didn't you see this?
My guess is that you didn't have a legacy of cranky IDEs from a
windows-sort-of-installation where you had to set the BIOS to NORMAL or LARGE to
get the disk to work at all. Disks, I might add, that will be rotated out as soon
as budget permits. All the other disks with this problem are also set either to
LARGE or NORMAL in the BIOS. The LBA mounts work sem to work fine.
I will report after this weekend whether this behavior persists over 7.1--I have to
shut down a server to test that.
But I think it is curious that you assume you are so correct that proper operation
at your site disproves my claim from experience at mine. Perhaps you wrote that
response faster than you intended?
Civileme
<SNIP>
> Regards,
>
> Ron. [AU] - sent by Mandrake Linux.
--
BETA testing KDE2 and Konqueror
