Replace error tracking with a way to monitor cache misses

What is the need / problem

We’re mixing both errors with cache misses in Bugsnag which makes it hard to have a clear on the actual errors while fails on providing a way to keep track of the products cache hit/miss rates.

Who does it impact

The sysadmin/dev team

What is the current impact of this problem

All instances are affected.

What is the benefit in focusing on this

We will have a way to track the hit/miss rate of the products cache and take decisions based on that and spot problems. Now there’s no way.

Also, it’ll make it easier to go through the current errors in Bugsnag.

Potential solutions that will solve this problem

We can use some sort of log aggregation to build meaningful visualizations on top of them but there might be other solutions out there.

Links to more details

Related to: Better logging for helping to resolve production issues
Existing GH Issue: https://github.com/openfoodfoundation/openfoodnetwork/issues/1949

Tech debt. Closing this as out of scope of the icebox process.

@sigmundpetersen @sauloperez @maikel @luisramos0 feel free to do with the idea what you will.

At Coopdevs are working on a ELK (elasticsearch, logstash and kibana) solution meant for one of our customers but to be used for all our projects, that we hope soon can be used by OFN as well.

1 Like