5 Levels of High Availability: from Multi-instance to Hybrid Cloud
summary
Titles states the scope but specifics decide: five concrete HA levels are defined and compared, with Hazelcast examples and measured latencies — 5 levels: Multi-instance, Multi zone, Multi region, Multi cloud, Hybrid cloud and one experiment shows 80 milliseconds per 10 000 entries on a single instance versus under 1 ms with synchronous backups across machines, multi‑zone keeps backups in different AZs via zone aware partition groups, multi‑region requires asynchronous WAN/geo replication and accepts eventual consistency and higher latency.
AI-written summary · about
chapters (11)
- 0:00High availability challenges
- 1:16About speaker and Hazelcast
- 2:40High availability levels
- 5:10Single instance deployment
- 7:36Multi-instance replication
- 15:46Multi-zone deployment
- 21:23Multi-region geo-replication
- 28:30Multi-cloud architecture
- 31:41Hybrid cloud deployment
- 34:30Complexity vs cost trade-offs
- 36:35Q&A
related talks
33mMany Layers of Availability
34mHigh Performance Cloud-Native Microservices
38m50 Shades of Multi-Cloud Infrastructure
43mLeveraging Multi-Cloud Clusters in Real-World
44mAvailability, Latency and Cost: Withstanding Outages
19mPreparing for disaster: going Multi-Region
41mAn Adventure in Multi-Cloud with Kubernetes
41mSeven patterns to increase application availability