Here switch, switch, switch…

When I started my new job a few weeks ago, one of the first things I noticed was a lack of consistency in the infrastructure. When it came to hardware and software, it seemed like no two items were alike.

For example, in our corporate office we had 9 switches – for about 135 network devices. Not only were there 9 switches, but no two were the same type. We had three 3Coms, three HPs, and three Dell. Only two of the HPs were worth keeping, an HP Procurve 1800-24G, and an HP ProCurve 2610-24/12PWR. The other switches I scrapped and replaced with two HP Procurve 2510-48s. Now everything is interconnected with fiber, and I am monitoring all the important ports with MRTG. All of this is a good thing to me. Especially since the old switches were very low end, and were interconnected in more of a mesh than in a chain. It isn’t an understatement to say our network performance has increased significantly.

So, on to the datacenter. We only have 3 switches down there. One is external, on our uplink to IO, and then we have 2 more HP Procurve 1800-24Gs, one in the DMZ and one in the internal network. They’re OK for the time being, so I just wanted to set them up so I could query them with SNMP via MRTG, and get pretty graphs for their usage like I do at the office.

Unfortunately, I wasn’t able to locate the switches on the network. I thought that was odd, but then I remembered our switches in the office used dynamic IPs until I changed them as well, so perhaps these didn’t HAVE IP addresses. I set DHCP up on one of the servers in the datacenter, and waited for my switches to appear. Two devices I didn’t know existed popped onto the network when DHCP activated, but not the switches.

Next I tried scanning the subnet with Look@Lan. That didn’t work, so next I tried Angry IP. Still nothing. I finally broke down and installed Wireshark on my monitoring server, and collected the network traffic for a little while. I finally captured some LLDP traffic tagged as coming from an HP device. Buried in the data was the management address – on an entirely different subnet. So the switches were statically assigned to IP addresses I couldn’t reach.

Armed with that knowledge, I remoted in, changed the IPs, enabled SNMP, and got down to the business of monitoring my traffic.

Just one more opportunity in the last 3 weeks to ask “Really? That was the best idea you could come up with?”

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s