Monitoring the Future: Insights from VictoriaMetrics with Aliaksandr Valialkin | Live On Production

"Live on Production" is a podcast series hosted by Chinmay Naik, CEO of One2N. The podcast takes you into the world of production systems at scale, drawing from One2N's extensive experience in the field. Each episode features a guest expert who shares their insights and experiences, providing a deep dive into the challenges and strategies of production scale.

In this inaugural episode, Chinmay Naik welcomes Aliaksandr Valialkin, Co-founder and CTO of VictoriaMetrics, to discuss the latest trends and innovations in monitoring and observability. They explore the challenges of scaling monitoring systems, the role of open-source solutions, and the future of monitoring in the cloud-native era.

Read VictoriaMetrics docs - https://docs.victoriametrics.com/


Introduction (00:00 - 02:00)
Why Victoria Metrics was Built (02:00 - 10:32)
-- Problems with Existing Monitoring Solutions (01:50 - 04:00)
Zabbix required custom scripts for collecting metrics, graphs, and alerts.
Prometheus struggled to process a large number of custom metrics.
-- ClickHouse Inspired the Design of Victoria Metrics (04:00 - 10:32)
ClickHouse provided high performance and efficiency for data storage and processing.
Victoria Metrics leverages similar ideas for achieving high performance and efficiency.
Open Source Journey (10:32 - 14:43)
-- Initial Idea as a Cloud Solution (10:32 - 11:22)
Alexander Valialkin initially wanted to create a cloud-based observability solution.
-- Open Sourcing Due to User Adoption Needs (11:22 - 14:43)
Users encouraged open sourcing to gain traction.
This decision helped Victoria Metrics grow organically.
Victoria Metrics for Beginners (14:43 - 19:22)
-- Designed for Efficiency (14:43 - 15:22)
More resource-efficient than Prometheus in terms of memory, disk space, and disk I/O.
-- Backwards Compatibility with Prometheus (15:22 - 16:12)
Can be used as a drop-in replacement for Prometheus with minimal configuration changes.
-- API Compatibility with Prometheus (16:12 - 19:22)
Leverages a similar querying API.
Victoria Metrics Community (19:22 - 29:12)
-- Benefits of Open Source (19:22 - 21:22)
User feature requests and bug reports help improve Victoria Metrics.
Increased code quality due to a wider range of use cases being discovered.
-- Challenges of Open Source (21:22 - 29:12)
Difficulty keeping up with reviewing all pull requests due to a high volume.
Prioritization of features becomes necessary to avoid feature bloat.
Build vs Buy for Observability (29:12 - 34:22)
-- Recommendation (29:12 - 31:22)
Start with a vendor solution for beginners due to easier setup and lower resource requirements.
-- Switching to In-house (31:22 - 34:22)
Consider switching to an in-house solution when vendor costs become too high.
Cost Considerations (34:22 - 36:22)
-- Absolute Cost Threshold (34:22 - 36:22)
Consider switching to an in-house solution when the observability vendor costs become a significant absolute cost.
Rapid Fire Questions (36:22 - 44:44)
-- Victoria Metrics vs ElasticSearch (36:22 - 37:22)
Victoria Metrics uses up to 30 times less memory and 10 times less storage space.
-- OpenTelemetry (37:22 - 38:22)
OpenTelemetry is still in the middle of evolution.
-- Upcoming Feature in Victoria Metrics (38:22 - 44:44)
Excited about the new querying capabilities in Victoria Logs.
Believes this query language will become a standard for querying logs.
Easy to use and powerful for complex log queries.








Music from #Uppbeat (free for Creators!):

https://uppbeat.io/t/kidcut/flamingo

License code: TBEVP2BNJ4YQYAC4