Founded in 2000, Trupanion is a leading provider of medical insurance for cats and dogs throughout North America, and the first provider licensed to do its own underwriting. Based in Seattle, the publicly-traded company employs over 400 people, with employee headcount increasing at least 40 percent annually over the last four years.
Adrian DuPre understands the challenges of dealing with customers nationwide on a 24x7 basis. As vice president of IT services at Trupanion, DuPre is responsible for supporting multiple business-critical applications, including a website, CRM system, and sales and claims processing applications. With offices in multiple locations and business done through a web portal, this means employees work around the clock.
In a 24x7, B2C environment, business never stops
"We require a robust system monitoring environment to ensure that we can respond to any system errors or outages - ideally before our users or customers are impacted," DuPre said.
The environment DuPre is responsible for monitoring consists of seven VM hosts at a managed collocation facility as well as three in-house hosts used for development and testing - more than 100 virtual servers in total. Their network consists of five sites tied together by an MPLS cloud - three corporate offices, a hosting location in Dallas and their main hosting facility in Santa Clara.
DuPre had been monitoring his environment with LogicMonitor, a hosted network monitoring solution, but found it to be inflexible and unreliable. It did not offer a wide enough range of sensor types, and had the additional downside of needing connectivity.
"We were looking for a solution that's reasonably-priced, for the SMB market, that's easy-to-implement but also flexible, with a wide array of monitoring options," he said.
Finding the right solution
In his search for a new network monitoring tool, DuPre had four parameters - the solution needed to be rich in functionality, a value in price, easy to set up and flexible. After doing his own research and getting several recommendations, he ultimately selected PRTG Network Monitor. His decision was based on the full-feature set being available at a lower price point and that the software required minimal effort to get meaningful information. Additionally, he was impressed with the amount of sensor types, which provided the flexibility his department needed to monitor a complex and growing environment.
"We were very impressed by the 'auto-discovery' option which allowed us to quickly establish a baseline set of sensors on important servers," he said. "This gave us a reliable first pass at configuration while we continued to "divide and conquer" our detailed application monitoring needs."
DuPre now uses PRTG to monitor all production infrastructure in detail, everything from servers, network links, routers, and even building monitors into applications. He is also using PRTG to monitor their website and bandwidth utilization across routers and switches, and soon, IT will also be able to look at the traffic breakdown across MPLS links.
"PRTG provides much needed insight into the health of our systems and applications, and gives us advance notice of potential issues. IT is informed of a problem by the monitoring system rather than by the end-user," he said.
As an IT manager in a regulated industry, DuPre also has to deal with specific compliance regulations. As part of compliance with Sarbanes-Oxley, Trupanion uses PRTG to monitor Windows event logs during backups. Originally, this process had to be done manually.
"We're using PRTG to monitor event logs and flag failures in batch and backup processes - this covers a regulatory need to monitor these processes and saves a lot of administrative time otherwise spent on manual log reviews."
Continued benefits with PRTG
Trupanion has experienced a number of benefits since implementing PRTG. DuPre has fully optimized CPU, memory and disk usage, and is able to monitor the health of links to teleworker locations. More importantly, there is a sense of relief that comes with knowing the health of production systems are monitored 24x7, and that notifications and alerts will reach IT well before end users notice issues. Additionally, he noted that support from Paessler has been efficient, although he hasn't needed it much.
"IT is informed of a problem by the monitoring system rather than by the end-user - we are often able to prevent downtime, or at least respond more quickly to an incident, than we would have otherwise," he said. "We haven't looked into quantifying benefits yet, but the general feeling is PRTG is well worth the money, and provides significant benefits in the areas of system reliability and staff work efficiency."