Let’s .explain() MongoDB Performance | Twitch Live Coding
Query performance can either be a constant headache or the unsung hero of any application. A well-tuned database will save you time, resources, and money - not to mention avoiding emergency escalations. Technical Services Engineer and "Query Doctor" Chris Harris and Developer Advocate Karen Huaulmé will demonstrate how to discover and optimize slow running queries. We will then teach you how to use the .explain() method and decipher explain plans to gain insight into your database operations and pick just the right indexes for your workloads.
Time Stamps:
0:00 Introductions and how Karen tricked Chris into being the chump for this "Stump the Chump" session
5:48 Preview of Chris' MongoDB .Live talk
7:31 Agenda
9:32 Quick Atlas tier overview
11:00 Proper indexing example - M10 vs M30 comparison and the "Query Targeting" graph
17:39 Things to consider when encountering poor performance
19:36 Tools in Atlas that will help with resolving performance issues
19:36 Real time performance panel
21:01 Pro tip regarding metric timing and the real time performance panel
22:22 Query Profiler
23:09 Performance Advisor
25:22 Why is an operation slow? Consider hardware for analytical workloads
27:56 The life-changing improvements of tuning queries: 99+% improvements from building indexes
30:19 What is an explain plan? Let's turn an inefficient collection scan into an efficient index scan
33:30 Using indexes to satisfy a sort request
36:46 The "Scan and Order" Atlas graph
37:53 The three Atlas charts that you should look at related to query efficiency and workload
39:04 Compound indexes: ESR guidance and why key ordering is important
41:21 The general structure of explain output, winningPlan and rejectedPlans sections
46:27 Question about wildcard indexes
52:04 Interpreting an explain plan for an aggregation
54:59 gowthamraj100 catches Chris being lazy - What does explain(1) mean?
1:01:24 High level information about querying sharded clusters
1:04:37 sergiocasao asks about query predicates and compound index usage. The helpful queryPlanner.parsedQuery field and the importance of the first key in the index definition
1:07:56 Summarizing the effect of the "ESR Rule"
1:11:25 gowthamraj100 asks about MongoDB's plan cache (don't forget about .pretty())
1:14:10 Is there such a thing as too many indexes? $indexStats is helpful for determining index usage
1:18:06 sergiocasao has a few additional questions regarding index usage and next steps. The Atlas UI itself allows you to perform many tasks related to indexing
1:24:08 The team wraps up with a special guest appearance!
Видео Let’s .explain() MongoDB Performance | Twitch Live Coding канала MongoDB
Time Stamps:
0:00 Introductions and how Karen tricked Chris into being the chump for this "Stump the Chump" session
5:48 Preview of Chris' MongoDB .Live talk
7:31 Agenda
9:32 Quick Atlas tier overview
11:00 Proper indexing example - M10 vs M30 comparison and the "Query Targeting" graph
17:39 Things to consider when encountering poor performance
19:36 Tools in Atlas that will help with resolving performance issues
19:36 Real time performance panel
21:01 Pro tip regarding metric timing and the real time performance panel
22:22 Query Profiler
23:09 Performance Advisor
25:22 Why is an operation slow? Consider hardware for analytical workloads
27:56 The life-changing improvements of tuning queries: 99+% improvements from building indexes
30:19 What is an explain plan? Let's turn an inefficient collection scan into an efficient index scan
33:30 Using indexes to satisfy a sort request
36:46 The "Scan and Order" Atlas graph
37:53 The three Atlas charts that you should look at related to query efficiency and workload
39:04 Compound indexes: ESR guidance and why key ordering is important
41:21 The general structure of explain output, winningPlan and rejectedPlans sections
46:27 Question about wildcard indexes
52:04 Interpreting an explain plan for an aggregation
54:59 gowthamraj100 catches Chris being lazy - What does explain(1) mean?
1:01:24 High level information about querying sharded clusters
1:04:37 sergiocasao asks about query predicates and compound index usage. The helpful queryPlanner.parsedQuery field and the importance of the first key in the index definition
1:07:56 Summarizing the effect of the "ESR Rule"
1:11:25 gowthamraj100 asks about MongoDB's plan cache (don't forget about .pretty())
1:14:10 Is there such a thing as too many indexes? $indexStats is helpful for determining index usage
1:18:06 sergiocasao has a few additional questions regarding index usage and next steps. The Atlas UI itself allows you to perform many tasks related to indexing
1:24:08 The team wraps up with a special guest appearance!
Видео Let’s .explain() MongoDB Performance | Twitch Live Coding канала MongoDB
Показать
Комментарии отсутствуют
Информация о видео
Другие видео канала
Data Resilience and Compliance StrategiesMongoDB and Accenture discuss the goals and challenges of cloud migrationForbes & MongoDB: Accelerating the Experience for Developers and End UsersA LATAM Success Story: PicapClaim Management Using LLMs and Vector Search for RAGGet More $out of AggregationFrom Connected Equipment to Connected Products (MongoDB for Manufacturing)Intro to Go - Part 2 - HTTP Servers Persisting Data in MongoDBBuilding AI Agents with MemoryAccelerating DevOps Agility: MongoDB Atlas in the Kubernetes PipelineManufacturing Digital Twin Data Modeling using MongoDBHow do I Build a Search Service for an App in Java with MongoDB?Data Model Your Way to Success with MongoDB Developer ToolsAzure Spring Apps and MongoDB Atlas First DeploymentThe Index: AI Vector Search and much more!!The Index: Ruby on Rails, FARM Stack Apps, & Semantic Search With OpenAIUnlocking Software Development Efficiency with UnqorkRevolutionizing Legal Work: How Genie AI and MongoDB Enhance In-House Legal TeamsThe Index: World Tour, RDBMS Schemas, & TerraformOptimizing Java Performance With Virtual Threads, Reactive Programming, and MongoDBLove Your Developers - No Matter the Future