We're All Distributed Systems Developers Now By Aaron Stannard, Founder & CEO Petabridge
High Expectations
The obvious way to scale is wrong.
Obvious Solution: Sharding
Brittle
Scenario 2: Real-time User Interactivity
Obvious Solution: Read-after-Write
Reality
Distributed Systems 101
Decentralization
Elastic (Join)
Elastic (Leave)
Recover from Failures
Availability through Replication
Event and Message Driven Programming
Properties of Messages
Messaging Patterns
Messaging Protocols
Gossip: How Nodes Discover Each Other
Stateful Application Programming
Stateful Apps Serve Results from Memory
Fastest Response Time?
State Locality
State Makes Protocols Work
CAP Theorem
CAP Terminologies
CAP Trade-offs
Highest Consistency?
Consistency vs. Availability
Fault and Resource Isolation with Microservices
WebCrawler Microservices
WebCrawler Network Topology
Try to make CPU / Memory- intensive tasks into stateless services
Stateful services should increase CPU / memory utilization slowly
Petabridge https://petabridge.com/
Recommend
More recommend
