| name | systems-computing |
| description | Understand computer systems from digital logic through operating systems, networks, databases, and distributed systems. |
| sasmp_version | 1.3.0 |
| bonded_agent | 05-systems-expert |
| bond_type | PRIMARY_BOND |
Systems Computing Skill
Skill Metadata
skill_config:
version: "1.0.0"
category: systems
prerequisites: [cs-foundations]
estimated_time: "10-12 weeks"
difficulty: intermediate-advanced
parameter_validation:
domain:
type: string
enum: [cpu, memory, os, network, database, distributed]
required: true
scale:
type: string
enum: [single, cluster, global]
default: single
retry_config:
max_attempts: 3
backoff_strategy: exponential
initial_delay_ms: 500
observability:
log_level: INFO
metrics: [domain_usage, design_pattern_frequency]
Quick Start
Master how modern computers actually work, from circuits to distributed systems.
Digital Logic & CPU
CPU Concepts
- Instruction execution cycle
- Registers and cache hierarchy
- Pipelining
- Branch prediction
Cache Memory
- L1, L2, L3 hierarchy
- Temporal and spatial locality
- Cache lines and blocks
Operating Systems
Process Management
- Process states and context switching
- Scheduling algorithms: FCFS, SJF, round-robin
Memory Management
- Virtual memory and paging
- Page tables and TLB
- Page replacement: LRU, optimal
Concurrency
- Race conditions and critical sections
- Mutual exclusion: locks, semaphores
- Deadlock: conditions, detection, recovery
Networking
OSI Model
- Layer 1: Physical (signals, cables)
- Layer 2: Data Link (MAC, switching)
- Layer 3: Network (IP, routing)
- Layer 4: Transport (TCP, UDP)
Protocols
- HTTP/HTTPS: Web
- TCP: Reliable connection
- UDP: Fast connectionless
- DNS: Name resolution
Distributed Systems
CAP Theorem
- Consistency: all nodes same data
- Availability: responds to requests
- Partition tolerance: survives splits
- Can guarantee only 2 of 3
Consensus Algorithms
- Paxos: proven consensus
- Raft: simpler consensus
Troubleshooting
| Issue | Root Cause | Resolution |
|---|
| High latency | Slow DB queries | Add indexes, cache |
| Memory exhaustion | Memory leak | Profile, fix leaks |
| Connection exhaustion | Pool too small | Increase pool size |
| Cascading failure | No circuit breaker | Add circuit breakers |
Key Concepts
- Latency: Time to complete one task
- Throughput: Tasks completed per unit time
- Scalability: Performance with increasing load
- Fault tolerance: Reliability despite failures
Interview Questions
- How do caches work?
- Explain virtual memory
- What is thrashing?
- Difference between TCP and UDP?
- How do transactions work?
- Design a system that scales to millions of users
