Keeping Argo CD sharp: Metrics, updates, and scaling insights
There’s more to scaling Argo CD than just throwing cpu and memory at the problem! This episode of Argo Unpacked is full of insights from Argo Maintainers: Dan and Alexandre about what a well running Argo instance looks like and what metrics admins should be tracking to prevent problems before they happen.
Here are some key insights along with their highlight video from the show:
Your Argo Instance Shouldn’t Constantly Be Reconciling!
Is Your Argo CD Reconciliation Doing Too Much
Argo CD has many scalable components but they may be scaling more than is necessary without proper configuration! In Argo CD reconciliation happens when a change is detected and it should be fast. Even if it takes 30 seconds to refresh, that could be a sign there is a problem.
Is your Argo CD running smoothly? Let the metrics decide
Monitoring Argo CD can go from tracking the entire sync journey to zooming in on specific metrics. Tracking key Argo CD metrics can help you catch slowdowns before they become problems. From the time it takes to fetch your Git repository to other steps in the reconciliation process, monitoring these details helps keep your setup smooth and reliable.
But remember, when you get rid of one bottleneck, it will usually expose others!
More from Argo Unpacked - Catch up on Argo CD 3.0 release
Argo CD 3.0 has been out for almost a month now and there have been 3 patch releases with minor fixes so if you haven’t upgraded, it may be time to start thinking about it.
Coming Up We Are Redefining the Multi-Cluster Story with Argo CD
Fellow Argo Maintainer Jann Fischer will be joining to discuss the work being done on Argo CD Agent, an open source way to extend Argo CD’s reach to clusters behind the firewall and offload capacity between more clusters.
