“Before we engaged Aster Labs, we had a baseline level of monitoring set up, but we kept seeing edge cases where something would go down or break that we hadn’t seen coming. Every time there was an outage, we would be scrambling to figure out what had happened, why things had happened that way, and what it was going to take to get diagnostics to avoid it for the next time. The holes we had in our monitoring and our lack of holistic strategy led to an outage that upset and enraged our non-technical business partners and eroded their trust in us.
In the San Francisco engineering culture, bringing in an outside expert to solve a strategy-level problem is difficult to justify when you have talented people working for you and when you consider technology strategy a core competency. But monitoring is a niche. You want to hire an expert when it comes to something that is as easy to get wrong as monitoring, and it’s the explicit area of expertise for Aster Labs. Mike Julian has a reputation in the monitoring space of being very effective in solving these problems in a short period of time. It really was a no-brainer to hire Mike.
Mike developed a strategy to help us approach monitoring less from the technical side and more from the perspective of the business metrics that mattered. With the strategy, when something went down that could affect the business, we had a detailed run book for how to deal with it and the potential causes for certain alert triggers. It gave us insight and visibility so we could then apply ourselves to problems that came up. We could identify problems before they caused user-facing problems.
From an external perspective, we went down less. Internally, it helped rebuild some of the eroded trust between engineering and the business. I could sleep better at night knowing that there wasn’t going to be something simple that would cause an outage. On the flip side, I also wouldn’t get woken up by the monitoring system at 3 a.m. for an alert that wasn’t really a problem. Having an intelligent approach to these things is important and that’s what Aster Labs provided us.
Everyone else I know who does monitoring for a living works for a vendor and the problem with that is that they will tell you why the tool they sell is the answer. Mike’s not wedded to a particular tool or implementation. He is focused on solving the problem and that solution will vary depending on who you are and what your constraints are.”