Running rancid for all of our cisco gear is pretty nice, but we’re acquiring more and more non-cisco (dell/F10) switches for our iSCSI 10Gb and 40Gb needs. Price is a major contributor to that shift.

Anyway, we needed to add Dell PowerConnect 8132 and 8132F switches to the monitor.

This is geared towards the Debian/Ubuntu install I did in an earlier post. Files are from GIT (https://github.com/dotwaffle/rancid-git/tree/master/bin)

Download the required bin files
cd /var/lib/rancid/bin
wget http://it.thelibrarie.com/utilities/dlogin
wget http://it.thelibrarie.com/utilities/drancid
chmod +x dlogin
chmod +x drancid

Edit the Rancid manufacturer file
nano rancid-fe
Under %vendortable add the Dell brand
'dell' => 'drancid',
Save and Exit

Open your router.db file
nano /var/lib/rancid/IMO/router.db

Set your switch in the database
your.dell.switch:dell:up
Save and Exit

Either run rancid or wait until your cron job kicks off
sudo -u rancid -H /usr/bin/rancid-run

****EDIT 4/2/14****
So I ran the dlogin/drancid against one of my Dell switches and was receiving the following error in my /var/log/rancid/newestfilehere:

exec failed router manufacturer dell: No such file or directory

Dang, that’s no good.

cd /var/lib/rancid/bin
./dlogin IPADDRESS_OR_NAME_OF_DEVICE

-bash: ./dlogin: @EXPECT_PATH@: bad interpreter: No such file or directory

Well that solves that issue.

nano /var/lib/rancid/bin/dlogin
Add this to the very top:
#!/usr/bin/env expect
Save and quit
./dlogin IPADDRESS_OR_NAME_OF_DEVICE

root@rancid:/var/lib/rancid/bin# ./dlogin fp-dell8132sw1
fp-dell8132sw1
spawn ssh -c 3des -x -l dellsw fp-dell8132sw1
The authenticity of host ‘fp-dell8132sw1 (10.20.209.000)’ can’t be established.
RSA key fingerprint is 0b:05:08:c1:00:e9:00:30:0d:00:00:00:00:00:00:00.
Are you sure you want to continue connecting (yes/no)?
Host fp-dell8132sw1 added to the list of known hosts.
yes
Warning: Permanently added ‘fp-dell8132sw1,10.20.209.000’ (RSA) to the list of known hosts.
dellsw@fp-dell8132sw1’s password:

FP-DELL8132SW1>enable
Password:********

Success!

And make sure you do the same to @PERLV_PATH@: bad interpreter: No such file or directory
nano /var/lib/rancid/bin/drancid
Add the following to the top of the file

#!/usr/bin/perl -w

Save and close
./drancid to verify

EDIT!!!
apparently my hosting company doesn’t like me having perl, python, or other scripts laying around in a directory with no index page. So here’s the tarball of the files:
Tarball of 3 files

Installed BYOBU (easier for me to use than screen) on my security-onion
(http://blog.securityonion.net/2012/02/byobu-and-security-onion.html)
apt-get install byobu

Execute byobu
byobu
F9 and set Byobu to launch automatically

Configuration (Pre-12.04 Ubuntu)
byobu-config
Toggle status notifications
Scroll down space bar to disable the logo
Tab to apply and then exit

Configuration (12.04 and later Ubuntu)
byobu-select-backend
Select Screen

Using BYOBU
Control + a = start of menu (F9 for more settings)
Control + a, c = create a new screen
Control + a + a = last screen used
Control + a, 0 = screen 0
Control + a, 3 = screen 3

On a development Oracle environment we were running Oracle (RHEL 6.x) on a Dell PowerEdge R510 server and connected via iSCSI to a SuperMicro NAS (Nas4Free 9.1.x) on a bonded 1Gb ethernet switch (LACP on Cisco). We then changed out the one SuperMicro NAS for an updated and larger NAS and reconnected to the Oracle environment. Unfortunately we also saw attempts to contact the old/retired NAS on iSCSI.

View all iSCSI sessions
iscsiadm -m session

tcp: [2] 10.10.0.137:3260,1 iqn.2013.us.temp.zfs.freenas1:disk0
tcp: [3] 10.10.0.138:3260,1 iqn.2014.us.temp.zfs.freenas2:disk0

Logout of an iSCSI connection
iscsiadm -m node --targetname "iqn.2014.us.temp.zfs.freenas2:disk0" --logout

Logging out of session successful

Discover all available iSCSI on a host
iscsiadm -m discovery -t st -p IPADDRESSOFHOST_ORGROUP

Login to all discovered iSCSI
iscsiadm -m node --loginall=all

Login to specific iSCSI target
iscsiadm -m node -T FQN_IQN:DISK_NUMBER -l -p IPADDRESSOFHOST_ORGROUP

I had a CentOS box that needed to backup files to another NAS system.

So I ran the following:
cp -rufT /source/dir/ /mnt/nas/dest/dir/
Everything copied without a hitch. I then edited one of the files and ran the copy command again. I was greeted with:

cp: overwrite ‘/mnt/nas/dest/dir/file’?

I forgot that CentOS, among other major providers, has an alias for the ‘cp’ command when running as root.

nano /root/.bashrc
Comment out the cp line
Rerun the copy command

/etc/rc.d/netif restart INTERFACENAME

I used to always write up iptables to block certain IP addresses, or use hosts.allow and hosts.deny entries, but Benjamin Cane showed me the light.

Blackhole that IP with a null route!

Add the route
ip route add blackhole IPADDRESSOFOFFENDER

Show the routes
ip route show

Remove the route
ip route del IPADDRESSOFOFFENDER
ip route show

Verify the null route
ssh: connect to host IP port 22: No route to host.