How to Monitor and Size Workloads on AWS i3 instances
2 likes557 views
The webinar focuses on monitoring and sizing workloads for ScyllaDB on AWS i3 instances, highlighting Scylla's performance capabilities such as over 1 million operations per node and under 1ms latency. Key topics include the use of monitoring tools like Prometheus and Grafana, data models, and operational impacts for effective performance management. The session also covers deployment strategies, insights on avoiding bottlenecks, and offers a Q&A for participant inquiries.
How to Monitor and Size Workloads on AWS i3 instances
- 1. Glauber Costa Eyal Gutkind ScyllaDB webinar How to Monitor and Size Workloads on AWS i3 Instances
- 2. + Intro: Who we are, what we do, who uses it + Getting started with Scylla + Using monitoring data to understand bottlenecks + Data model and operation impacts + Scylla compatibility + Q&A Today we will cover:
- 3. What we do: Scylla, towards the best NoSQL + > 1 million OPS per node + < 1ms 99% latency + Auto tuned + Scale up and out + Open source + Large community (piggyback on Cassandra) + Blends in the ecosystem: Spark, Presto, time series, search ...
- 4. Where Scylla is deployed?
- 5. Scylla benchmark by Samsung op/s
- 6. + Intro: Who we are, what we do, who uses it + Using monitoring data to understand bottlenecks + Why AWS i3 Instances and Scylla? + Data model and operation impacts + Scylla compatibility + Q&A Today we will cover:
- 7. Monitoring: Prometheus + Grafana - Container based installation - Simple and clean - https://github.com/scylladb/scylla-grafana-monitoring - Installs on a different machine - Possible to install w/o docker - Bring your own Prometheus and Grafana - Use provided yml, ini and json files to enable monitoring
- 8. How to Understand Scylla Performance
- 9. What to look for? Customize for workload
- 10. Setting up the monitor
- 11. + Intro: Who we are, what we do, who uses it + Using monitoring data to understand bottlenecks + Why AWS i3 Instances and Scylla? + Data model and operation impacts + Scylla compatibility + Q&A Today we will cover:
- 12. Powerful software for powerful instances
- 13. Threads Shards System got muscle?, use it! : Shard per core
- 14. Powerful software for powerful instances
- 15. Don’t leave RAM behind: Unified cache Cassandra Scylla Key cache Row cache On-heap / Off-heap Linux page cache SSTables Unified cache SSTables Complex Tuning
- 16. Unified cache vs Cassandra cache plethora Cassandra Scylla Key cache Row cache On-heap / Off-heap Linux page cache SSTables Unified cache SSTables SSTable page (4k) Your data (300b) Parasitic Rows - 300b/4kb
- 17. Cassandra Scylla Key cache Row cache On-heap / Off-heap Linux page cache SSTables Unified cache SSTables App thread Kernel SSD Page fault Suspend thread Initiate I/O Context switch I/O completes Interrupt Context switch Map page Resume thread Page fault Unified cache vs Cassandra cache plethora
- 18. Show CPU and memory workload on the monitor
- 19. Powerful software for powerful instances
- 20. Exploiting NVMe performance: Scylla I/O Scheduling Query Commitlog Compaction Queue Queue Queue Userspace I/O Scheduler Disk Max useful disk concurrency I/O queued in FS/deviceNo queues
- 21. Show I/O workload on the monitor
- 22. Powerful software for powerful instances
- 23. When to use i3 instances
- 24. + Intro: Who we are, what we do, who uses it + Getting started with Scylla + Using monitoring data to understand bottlenecks + Data model and operation impacts + Scylla compatibility + What’s Next? + Q&A Today we will cover:
- 25. Prevent Hot Spots CREATE TABLE sensordata (s1data text,s2data text,timestmp timestamp,sregion text,stype text, PRIMARY KEY (s1data, s2data,timestmp)); CREATE TABLE sensordata (s1data text,s2data text,timestmp timestamp,sregion text,stype text, PRIMARY KEY ((s1data, timestmp), s2data));
- 26. Partitions size
- 27. + Intro: Who we are, what we do, who uses it + Getting started with Scylla + Using monitoring data to understand bottlenecks + Data model and operation impacts + Scylla compatibility + Let’s show you how to do it (Demo time) + Q&A Today we will cover:
- 28. + SSTable file format + Configuration file format + CQL language + CQL native protocol + JMX management protocol + Management command line Compatibility
- 29. + Intro: Who we are, what we do, who uses it + Getting started with Scylla + Using monitoring data to understand bottlenecks + Data model and operation impacts + Scylla compatibility + Q&A Today we will cover:
- 31. NoSQL GOES NATIVE Thank you.