SREDAY

Site Reliability, DevOps and Cloud

Nov 8, 2024 San Francisco, CA, USA

1
Days
16+
Speakers
1
Tracks
100
Attendees

New approach to metrics observability for the cloud era: cut costs and complexity with serverless and S3

Kiran Gollu
Oodle.ai

In this talk, we will talk about lessons learnt from taking a new first principles approach to metrics observability to achieve order of magnitude cost-efficiency and simplicity. Through eliminating disks and egress costs completely with AZ-aware architecture, you can truly achieve limitless scale and zero management overhead. By making an infrastructure primitive 10x cheaper, we can truly unlock the new use cases of metrics observability. The columnar based storage engine architecture allows you to send any cardinality (customer-id, request-id, user-id, pod-id, camera-id etc). Most observability tools downsample metrics (for 1-min or 5 min or 1 hour) depending on how the old data is. In this talk, we will cover how we can retain high fidelity data all the time with no downsampling at all while being cost-efficient.

Kiran Gollu is the Co-founder/CEO of Oodle.ai. Previously, he was head of cloud platform engineering at Rubrik (NYSE: RBRK), when he was responsible for observability, distributed databases, reliability/costs engineering and developer experience while they were scaling from <5M to 600M in ARR. He was also co-founder of YC backed observability startup, and was an early engineer of AWS S3 and DynamoDB.

Sponsors & Partners

Want to become a sponsor? Get in touch!