Issues with HD space.

[x86i]

Afternoon everyone,

I am surprised I haven't noticed this problem on my debian box before, but the / partition was made rather small when the server was setup.

I haven't had a need to increase it until I started trying to update to a new kernel. Now I am out of space on that partition.

df -h
Filesystem            Size  Used Avail Use% Mounted on
/dev/hda1             259M  256M     0 100% /
tmpfs                 507M     0  507M   0% /lib/init/rw
udev                   10M   60K   10M   1% /dev
tmpfs                 507M     0  507M   0% /dev/shm
/dev/hdb1              74G  1.5G   69G   3% /backup
/dev/hda9              63G  1.9G   58G   4% /home
/dev/hda8             373M   11M  343M   3% /tmp
/dev/hda5             4.6G  1.4G  3.0G  33% /usr
/dev/hda6             2.8G  208M  2.5G   8% /var

Now, I guess my question is. Can I resize that partition, taking some away from /home? Say resize /home by like 5gigs, and allocate that space to /?

Since I do not have physical access to the box I cannot simply use a live CD to unmount /.

Any advice?

Thanks,

X

[wgiese]

You're pretty much hosed, given you're remote. As you're probably aware, it's not a trivial task to do what you want with physical access to the server.

Assuming you have jag deal with things, you may wish to address the other problem with your setup at the same time - Jag's default IDE hd config is not ideal. I ran into it on my dedicated box when first set up, fortunately I caught it up front.

In a dual drive IDE setup, they put both drives on the primary IDE controller - one master, one slave. In an IDE environment, this hurts performance when accessing from both, and really hurts when writing from one drive to the other. A far better setup, is to have them move your second drive to hdc (secondary IDE, master).

Personally, I'm not in favor of using the second drive for backup - which given their default mount point, seems to be what they anticipate it being used for. On my debian dual-drive boxes, I'm more interested in redundancy - I have them RAID1 the drives, and save backups locally using automysqlbackup on the server, and a cron'ed rdiff-backup from my local box. Totally automatic..set it and forget it.

rdiff-backup is available in the debian repositories, automysqlback is just a script - google if interested.

[x86i]

Thanks for the reply. I pretty much assumed the same, and am having the OS reloaded with better partitioning. Luckily, I hadn't started to deploy anything just yet so I dodged that bullet.