Ljubljana, Slovenia

Api

SQL as an API
If you haven’t been living under a rock, you’ll know that recently there is an uptake in popularity of “Functions as a service”. In the open source community, the OpenFaaS project from Alex Ellis has received good traction, and recently Amazon Lambda announced support for Go. These systems allow you to scale with demand, and execute your CLI programs via API. Motivation behind Lambda/FaaS Let’s call this for what it is - the whole “serverless” movement is marketing for cloud stacks like AWS, that allow you to hand over any server management to them, for, hopefully, a fraction of your income.
Read more

So far there is a distinct lack of snakes
In the dashboard which I hacked together at MMC RTV Slovenia, I display a set of related flickr searches based on the current state of the system. So far, there's a distinct lack of snakes, volcanoes, lava, fire, explosions and other things that might freak an engineer out as much as a server outage. Thoughts: I like how the hosts are clearly marked when one of their metrics goes over a threshold.
Read more

Flickr API
I’m currently applying for a Flickr API key, and this is what I put into the description for “HQ Dashboard”: “I’m building a monitoring dashboard which includes statistics data from internal sources, some tweets and, hopefully, an assortment of Flickr backgrounds based on current monitoring conditions (ie., search for fire/destruction/storm/tornado when servers are being overloaded, and grass/meadows/happy/coffee if the monitoring says everything is fine). In this way I plan to use the Flickr search to provide a visual representation of the severity level currently being experienced.
Read more

API Backpressure
It's days like this where I love my job. I'm implementing back pressure for API calls that communicate to an external service. Depending on how overloaded the external service is, my API interface adapts to use caching more extensively, or to use a service friendly request retry strategy, minimizing impact on infrastructure and possibly even resolving problems when they occur. This is done by keeping track of a lot of data - timeouts, error responses, request duration, ratios between failed and successful requests,.
Read more