Tag: design-concurrency
All the articles with the tag "design-concurrency".
-
System Design - How the Linux Process Scheduler (CFS) Works
How Linux's Completely Fair Scheduler (CFS) works: virtual runtime, red-black tree scheduling, nice values, CPU bandwidth control, and source code walkthrough from the kernel's sched/fair.c.
-
System Design - How Two-Phase Commit (2PC) Works
How two-phase commit (2PC) works in distributed databases: the coordinator/participant model, prepare and commit phases, failure handling, recovery via write-ahead logging, and real-world implementations in PostgreSQL, MySQL, and TiDB.
-
System Design - How Linux epoll Works
How Linux epoll works: the evolution from select/poll, epoll's internal architecture with red-black trees and ready lists, level-triggered vs edge-triggered modes, and a source code walkthrough from the Linux kernel.
-
System Design - How Rate Limiting Works
How rate limiting works: token bucket, leaky bucket, fixed window counter, and sliding window log/counter algorithms with ASCII diagrams, source code walkthrough from Nginx, Guava, and Redis, and practical trade-offs.
-
System Design - How Apache Kafka Works
How Apache Kafka works internally: the distributed commit log, partitions, replication with ISR, producer batching, consumer groups, zero-copy transfer, and source code walkthrough from the apache/kafka repository.
-
System Design - How Garbage Collection Works
How garbage collection works in practice: reference counting vs. tracing, mark-and-sweep fundamentals, tri-color marking, Go's concurrent collector, write barriers, GC pacing, and source code walkthrough from the Go runtime.