After having been wanting to for ages I have finally fixed up my OpenBSD machines partitions.
I had an 80GB hard drive formated (wd1a) and moved /usr/local/ on to it and put my SMB shares on it for the free space.
Since wd0 is a 8GB disk split into a, b, h, d, g, and e partitions the biggest is wd0g mounted on /usr with ~6GB free but I had almost 10GB of files on /usr/local (wd1a). So I had to copy my backups and videos to the windows machine via Samba/Network Neighborehood before I could move all of my files in /usr/local/srv to a temporary place in /usr and then archived the rest of the directory.
cd /usr
mkdir storage
mv local/srv storage
tar -cf /var/tmp/local.tar local
I had to relabel the disk and then format the partitions, I created wd1a and wd1d to use as /usr/local and /srv with ~15GB more free space in case I need it.
umount -f local
disklabel -E wd1
newfs wd1a
newfs wd1d
During the disk label I changed to ‘disk geometry’ (g d), deleted the a partition (d a) and created the a and d partitions (c a and c d) keeping with the prompts on it and specifying 12G and 45G for the partition sizes.
Fixed my fstab and then mounted the partitions
vi /etc/fstab
# 8GB Primary Master, PATA drive
# device mount type opts dump fsck
/dev/wd0a / ffs rw 1 1
/dev/wd0h /home ffs rw,nodev,nosuid 1 2
/dev/wd0d /tmp ffs rw,nodev,nosuid 1 2
/dev/wd0g /usr ffs rw,nodev 1 2
/dev/wd0e /var ffs rw,nodev,nosuid 1 2
# 80GB Primary Slave, PATA drive
# device mount type opts dump fsck
/dev/wd1a /usr/local ffs rw,nodev 1 2
/dev/wd1d /srv ffs rw,nodev,nosuid 1 2
mount -o rw,nodev /dev/wd1a /usr/local
mount -o rw,nodev,nosuid /srv
I’m some what tempted to mark wd1d ‘noexec’ but I may wish to run scripts from there later if I ever move ~/code over. After that it was just a quick hop, skip, and jump to restore my files.
tar xpf /var/tmp/local.tar
mv storage/srv/smb /srv/
vi /etc/samba/smb.conf
I corrected all of my shares in smb.conf from command line mode:
:1,$s//usr/local/srv//srv/g
I could’ve used ex but I rather like paging up/down with ^U and ^D instead of using ‘addr1,addr2p’ in ex.
mount
/dev/wd0a on / type ffs (local)
/dev/wd0h on /home type ffs (local, nodev, nosuid)
/dev/wd0d on /tmp type ffs (local, nodev, nosuid)
/dev/wd0g on /usr type ffs (local, nodev)
/dev/wd0e on /var type ffs (local, nodev, nosuid)
/dev/wd1a on /usr/local type ffs (local, nodev)
/dev/wd1d on /srv type ffs (local, nodev, nosuid)
# df -h
Filesystem Size Used Avail Capacity Mounted on
/dev/wd0a 147M 30.4M 110M 22% /
/dev/wd0h 393M 35.6M 337M 10% /home
/dev/wd0d 98.3M 2.0K 93.4M 0% /tmp
/dev/wd0g 6.7G 398M 6.0G 6% /usr
/dev/wd0e 148M 84.1M 56.2M 60% /var
/dev/wd1a 11.8G 76.9M 11.1G 1% /usr/local
/dev/wd1d 44.3G 5.1G 37.0G 12% /srv
Windows wouldn’t see the file shares and sending the HUP signal to Samba to reread it’s conf file immediately didn’t help any. So I gave Vectra a reboot to double check my fstab entry (yes I am paranoid), I could’ve just killed the processes and reloaded them manually for the same effect.
# uptime
9:14PM up 19 days, 3:29, 1 user, load averages: 4.12, 4.16, 3.86
# reboot
I love OpenBSD 🙂
EDIT:
To prevent some nasty time outs.
vi /etc/ssh/sshd_config
ClientAliveInterval 15
ClientAliveCountMax 45
vi ~/.ssh/config # or /etc/ssh/ssh_config for all clients
ServerAliveInterval 15