Successfully reported this slideshow.
We use your LinkedIn profile and activity data to personalize ads and to show you more relevant ads. You can change your ad preferences anytime.

Couchbase 5.0 at the speed of memory: operating 100% in memory – Couchbase Connect New York 2017

443 views

Published on

In their latest In-Memory Database Wave, Forrester says, “An in-memory database is not just a nice-to-have option anymore – it has become critical to support next-generation transactions, analytics, and operational insights.”
This session will provide a deep dive into the newest in-memory database features included in Couchbase Server, in particular ephemeral buckets. Additionally, we will compare and contrast standard global secondary indexes with memory-optimized indexes (MOI). You will learn how ephemeral buckets can be leveraged with MOI for high speed in-memory database operations without requiring any data persistence.

Published in: Software
  • Be the first to comment

  • Be the first to like this

Couchbase 5.0 at the speed of memory: operating 100% in memory – Couchbase Connect New York 2017

  1. 1. ©2017 Couchbase Inc. Operating 100% in Memory Ephemeral Buckets & Memory-Optimized Indexes 1
  2. 2. ©2017 Couchbase Inc. Agenda • In-memory databases • A brief history of buckets • Introducing Ephemeral Buckets • Memory Optimized Indexing (MOI) • Demo 2
  3. 3. ©2017 Couchbase Inc.©2017 Couchbase Inc. 3 In-Memory Architectures
  4. 4. ©2017 Couchbase Inc. “An in-memory database is not just a nice-to-have option anymore - it has become critical to support next-generation transactions, analytics, and operational insights.” - Forrester Wave™: In-Memory Databases, Q1 2017 4 In-Memory Databases
  5. 5. ©2017 Couchbase Inc. 5 In-Memory Database Requirements • Application • High ingestion rate • Low indexing latency • Complex query support • High query throughput • Connectors • Operations • High scalability • High availability • Ease of operations • Multi-model
  6. 6. ©2017 Couchbase Inc.©2017 Couchbase Inc. 6 A Brief History of Buckets
  7. 7. ©2017 Couchbase Inc. A Brief History of Buckets 7 Long, long ago
  8. 8. ©2017 Couchbase Inc. A Brief History of Buckets 8 Long, long ago 2010
  9. 9. ©2017 Couchbase Inc. A Brief History of Buckets 9 Long, long ago 2010 2012
  10. 10. ©2017 Couchbase Inc. Memcached vs Couchbase Buckets 10 Memcached Bucket Couchbase Bucket Persistence ✗ ✔ Replication ✗ ✔ Rebalance ✗ ✔ XDCR ✗ ✔ N1QL ✗ ✔ Indexing ✗ ✔ Max Object Size 1MB 20MB
  11. 11. ©2017 Couchbase Inc. Performance 11
  12. 12. ©2017 Couchbase Inc. But I don’t need persistence! • Cache • Sessions • Shopping carts • Analytics 12
  13. 13. ©2017 Couchbase Inc.©2017 Couchbase Inc. 13 Introducing Ephemeral Buckets New in Couchbase 5.0
  14. 14. ©2017 Couchbase Inc. Bucket Comparison 14 * MOI, FTS only Memcached Couchbase Ephemeral Persistence ✗ ✔ ✗ Replication ✗ ✔ ✔ Rebalance ✗ ✔ ✔ XDCR ✗ ✔ ✔ N1QL ✗ ✔ ✔ Indexing ✗ ✔ ✔* Max Object Size 1MB 20MB 20MB
  15. 15. ©2017 Couchbase Inc. Ephemeral Bucket Benefits • No high performance disk subsystem required • Lower costVMs • Smaller chassis • Even more consistent high performance • No disk IO contention (i.e. compaction) • Lower CPU consumption • No DiskWrite Queue • No IO threads • Faster maintenance operations • No warm-up • Faster node restart • Faster rebalance – currently 4x faster in our lab! 15
  16. 16. ©2017 Couchbase Inc. Ephemeral Bucket Limitations • Data set must fit in memory • Configurable OOM handling • No automatic recovery from total power loss • Backups and XDCR still supported! • Only Memory Optimized Indexes (MOI) and FullText Search (FTS) are supported • NoViews, GSI 16
  17. 17. ©2017 Couchbase Inc.©2017 Couchbase Inc. 17 Memory Optimized Indexes (MOI) Released in Couchbase 4.5
  18. 18. ©2017 Couchbase Inc. >50xHigher Indexing Rate 0 20000 40000 60000 80000 100000 120000 140000 Queries/sec Memory Optimized GSI Standard GSI Memory Optimized Indexes (MOI)
  19. 19. ©2017 Couchbase Inc. >20xHigher QueryThroughput (background workload: 30K updates/sec) 0 2000 4000 6000 8000 10000 12000 14000 16000 Queries/sec Memory Optimized GSI Standard GSI Details: Query Throughput (queries/sec), N1QL 1 bucket 20M items with 1K size, Query: Singleton Unique Lookup with stale=false, Mutations: 30KSops, Index: MOI, HW: 6 nodes x 24 cores 128GB RAM – 4 data, 1 index and 1 query service node Memory Optimized Indexes (MOI)
  20. 20. ©2017 Couchbase Inc.©2017 Couchbase Inc. 20 Demo
  21. 21. ©2017 Couchbase Inc. 22 The CouchbaseConnect mobile app Take our in-app survey!
  22. 22. ©2017 Couchbase Inc. 23 Share your opinion on Couchbase 1. Go here: http://gtnr.it/2eRxYWn 2. Create a profile 3. Provide feedback (~15 minutes)
  23. 23. ©2017 Couchbase Inc.©2017 Couchbase Inc. Follow us on social media Twitter @couchbase Facebook /couchbase Instagram @couchbase Linkedin /company/couchbase

×