Have you ever found yourself frantically trying to un-commit some enormous binary file you’ve accidentally checked in to version control? Or perhaps you’ve made the mistake of committing your AWS root credentials? Enter the BFG.
We have a PostgreSQL database server that stores a bunch of product catalogs. It weighs in at around 2 TB. The vast majority of the time, this data is just sitting there on SSD provisioned storage, waiting to be accessed by a batch indexing job.
GitLab has a lovely feature that can bring up a Kubernetes cluster and install build runners on it with a single click. But you can’t tweak it.
This one took me a while to figure out.
What began as an investigation into a very stubborn intermittent error turned into a write-up on how to squeeze the most out of your Kubernetes deployment.