Understanding and tuning WiredTiger

on YouTube ↗46m

" summary

Covers the WiredTiger storage engine introduced in MongoDB 3.0, explaining its architecture (document-level MVCC locking), compression support, and the option between B-tree and LSM indexes. Describes how to tune and monitor WiredTiger for improved performance.

AI-written summary · about

" chapters (21)

1. 0:00Start
2. 1:06Most popular NoSQL database
3. 1:235 NoSQL categories
4. 2:26MongoDB is a Document Database
5. 3:48Operational Database Landscape
6. 11:40MongoDB 3.0 Storage Engine API
7. 14:11Some existing engines
8. 16:18What is Wired Tiger
9. 19:54Choosing Wired Tiger at server startup
10. 20:26Main tunables exposed as MongoDB options
11. 21:41All Wired Tiger options via configString (hidden)
12. 23:45Understanding Wired Tiger architecture
13. 26:22Covering 90% of your optimization needs
14. 27:04fit working set in Cache
15. 27:33fit working set in OS Disk Cache
16. 29:16Strategy 2
17. 29:48SSD disk + compression to save
18. 30:16SSD disk (no compression)
19. 30:44Compression benchmarks
20. 33:11What problem is solved by LSM indexes?
21. 36:052B inserts (with 3 secondary indexes)

" related talks

• 01h 03mWiredTigerKeith Bostic · The Databaseology Lectures (Fall 2015)
• 28mComparing MongoRocks, WiredTiger and mmapv1 for performance and efficiencyMark Callaghan · Percona Live Open Source Database Conference 2017
• 19mMongoDB Engines: DemystifiedSveta Smirnova · Percona Live Open Source Database Conference 2017
• 20mMongoDB size managementSteven Wang · Percona Live Open Source Database Conference 2017
• 47mSchema design for fast applicationsRick Houlihan · AWS re:Invent 2023
• 51m5 boring performance improvements in MongoDB 3.4 that matterHenrik Ingo · HighLoad++ 2017
• 25mPercona Server for MongoDB: Features OverviewJon Tobin · Percona Live Open Source Database Conference 2017
• 41mWant to turbocharge your database? Ask me how!Danielle Monteiro · Women Dev Summit 2018