ipHouse is redefining hosting by providing businesses with state of the art virtual data centers. Instead of companies having to go out and buy hardware, put it into a rack and pay for colo fees and power, ipHouse provisions a virtual data center complete with CPU, RAM, disk space and network. The flexible architecture gives companies an easy growth path, yet much more consistent pricing than having to pay for each new instance as in Amazon.
Why is monitoring important to ipHouse?
Everything fails at some point or another. Whether it’ s a hard drive, a server going belly up, or an Apache process going slow, we need to know that before our customer notices or they get pissed, and rightfully so.
Prior to LogicMonitor, how were you monitoring your infrastructure?
We used Nagios as the monitoring engine, and Cacti for measurement. The two are different and don’t integrate well without a lot of hackery. They are built around the idea that you have someone who’s willing to waste time every day constantly tuning and changing things. I’m not paying someone for that.
How did you come across LogicMonitor?
I went to VMworld in part looking for products that do monitoring and measurement. There were a bunch of companies that were doing stuff just for VMware, but that’s only one piece of it. I was looking for one tool that does all of it.
Shortly after I got back, I stumbled across LogicMonitor on Google. LogicMonitor appeared to do the full stack, so we decided to test it.
How did you find the process of getting LogicMonitor up and running?
I can tell you that we tasked our entire staff – including sales and marketing people – with adding their own hosts into LogicMonitor, deploying their own agent, and learning how the system worked. The less technical people had a few issues, but in the end everyone thought it was great. That’s not an exercise we could have done with any other monitoring tool.
What are you monitoring with LogicMonitor?
Every type of gear that we have is currently in LogicMonitor — Juniper routers, Cisco switches, firewalls, modem pools, F5 load balancers, VMware servers, NetApp storage…We now monitor all of that in a single pane of glass, so I don’t have to watch a bunch of different web pages to make sure our services are running. I can see everything in one interface.
That integration is nice during remote times. If it’s 3:15 in the morning and my phone goes off, I don’t have to figure out which interface just paged me. I just go to one place for everything.
How have you found LogicMonitor’s depth of data?
LogicMonitor crushes other monitoring tools in the number of devices it supports out of the box, and the amount of datapoints it pulls out for you. It gives you more data than you could ever need. I’ve never had that before – it’s kind of nice to be in that realm for a change.
The data is also contextual. It gives you hints on how better to tune stuff. For example, on our MySQL Servers, in addition to sending triggers, it might say “Warning, query cache is being purged for low memory yet your query cache is close to 100%. You may want to add some more query cache.”
That’s a small thing, but if you get enough small stuff going at once, now it’s a big deal. This is the first monitoring tool that actually got us excited.
How has LogicMonitor’s support been?
The tech guys have been great to deal with. We work with them like a partner instead of the normal vendor-support thing where it’s:
Step 1: “You’re wrong.”
Step 2: “You’re still wrong.”
Step 3: “Maybe there is something here.”
We’re not getting that from the tech folks at LogicMonitor. We’re able to reach step 3 immediately and have a discussion back and forth. It could very well end up being something on our side. I really like that the normal pushback of “it’s all your fault” doesn’t happen with LogicMonitor. It is very refreshing and something every company should be able to expect from their vendors.
