Kibana and Grafana are written in AngularJS and AngularJS uses dirty watch for two way data binding. As performance issues can easily come up when the number of table data is more due to dirty watch, why would AnglarJS have been chosen as the preferred framework?

My guess is that they chose Angular because 1) client-only is very easy to deploy (no install, just download a folder) and 2) the “directive “ approach makes it very modular, which fits the need of a tailor-made dashboard.

The 2-ways data binding is an issue only if you load a lot of data on the client, which doesn ‘t happen because Kibana requests are written using things like count, facets and now aggregations. Elasticsearch actually does the data-intensive part and Kibana just need a neat way to inject graph representations in specific parts of the dashboard ‘s DOM. Angular is good at this.

This text was originally published in quora.