Network Flow Aggregation: Pay for the logs you care about!

Hubble flow logs are a lifesaver for debugging and security. Until you look at the cloud storage bill.
Millions of near-identical flows pile up, turning observability into an expensive 'nice-to-have' when it should be a 'must-have'!
We’ve had ways to trim logs before by using FieldMask and Include/Exclude filters or using set times to make flows more relevant and less costly, but those only go so far.
Our new approach - Flow Aggregation. Instead of storing each flow, we aggregate in-agent based on fields user actually cares about.
The result? Lower storage overhead, quicker troubleshooting, and leaner data that’s ideal for analytics and machine learning.
And yes, Calico users have had this for a while, now Cilium does too.
In this talk, we’ll walk through this new feature and why it matters, show a live demo comparing raw and aggregated flows, and share practical tips for tuning aggregation to your workloads.

If you love observability but hate runaway costs, this session is for you.