Opsee – Continuously test your services and deploy with confidence | Hacker News Reader

Opsee – Continuously test your services and deploy with confidence

opsee.com

72 pointsbtmerr14 comments

Noticed this typo:

> Built for your your stack

I do not have a valuable comment outside of that. I wish I could play with opsee, but I don't have anything going on right now that's big enough :(

Thanks! Fixed.

There are lots of products in this space. Pingdom is the gorilla, and there are lots off smaller companies & free services.

What makes this one different?

Co-founder here. The big difference, compared with tools like Pingdom, is that we're operating inside your environment. We can tell you where the failure happened and give you tools to fix it. We're like Pingdom for microservice environments, helping developers know their services, inside or outside the firewall, are working as expected.

Would you say you're competing with Nagios/Zabbix/Sensu/OpsServer?

Yes, definitely. That will become more clear when we launch support for RDS/SQL health checks in the next few weeks. The biggest differentiator between us and these OSS tools is ease-of-use and maintenance. With Opsee there are no configurations to manage, YAML files to create, or agents to run. Setup is dead simple, and you never need to maintain health checks. Our checks react instantly to changes in your infrastructure.

I saw a demo of this a few months ago, and it looks like they're reframing the product now with Health Checks being the primary feature. I know it's a small difference from monitoring, but I'd be interested to know why the change in framing.

It seems like everyone conflates the term monitoring with metrics, so instead of spending a lot of words in our messaging trying to redefine it, we felt it's more straight foward to frame everything around health checks. Hope that answers your question.

Free during public beta... and then what? Rather just know the pricing model outright.

Hey, I'm one of the co-founders of Opsee. Part of our beta is collecting enough data to figure out a reasonable pricing structure. We're leaning pretty heavily towards charging based on the number of health checks you have setup, independent of the number of instances you're checking. So it'll be much closer to how pingdom does pricing than a new relic, for instance.

Thanks. New Relic incentivizes running fewer, larger instances, which unnecessarily exposes to risk of machine failure.

Exactly, and with a containerized infrastructure the per host pricing model breaks down further. Our goal is to have the most frictionless pricing model for people with a microservice architecture.

Just to be clear: we (New Relic) explicitly state on our pricing page that we do NOT charge for containers. It's in our FAQ. We view them as processes, not hosts.

I saw a first-hand demo of this a couple of months ago - looks like a great way to start external monitoring without putting in too much effort, and will easily cover the needs of many public-facing sites and services.

Reply on news.ycombinator.com