distributed-consensus-reliability

What it does and when to use it

Reason about distributed consensus for reliability: when to use Paxos/Raft style systems, failure modes, and operational risks from Google SRE. Use when designing strongly consistent control planes or debugging quorum pain.

How it works

1. When to use
2. Principles
3. Workflow
4. Decision guide
5. Output template
6. Anti-patterns
7. Related skills
8. Canonical references
9. Agent notes

---
name: distributed-consensus-reliability
description: >-
  Reason about distributed consensus for reliability: when to use Paxos/Raft style systems, failure modes, and operational risks from Google SRE. Use when designing strongly consistent control planes or debugging quorum pain.
---

# Distributed Consensus Reliability

**Domain:** `resilience`  
**Sources:** SRE Book 1e Ch23

## When to use

- Choosing consensus-backed storage/coordination
- Quorum outages / leader flaps
- Control-plane reliability reviews

## Principles

- Consensus is powerful and operationally expensive — use where consistency is required.
- Quorum membership, latency, and network partitions dominate reliability.
- Separate control plane consensus from bulk data path when possible.
- Observe leader changes, commit latency, and peer health.

## Workflow

1. Confirm consistency requirement vs eventual OK.
2. Size quorum and failure tolerance (e.g. 2f+1).
3. Place members across failure domains carefully.
4. Define backup/restore and membership-change runbooks.
5. Alert on quorum risk, not only process down.

## Decision guide

- If cross-region sync latency breaks SLOs → rethink topology or consistency.

## Output template

```markdown
# Consensus reliability review
- Why consensus:
- Topology / failure tolerance:
- Observed risks:
- Runbooks:
```

## Anti-patterns

- Using ZooKeeper/etcd for huge data.
- Single-region quorum for multi-region hard needs.

## Related skills

`data-integrity`, `designing-for-resilience`

## Canonical references

- [Distributed Consensus](https://sre.google/sre-book/managing-critical-state/)

## Agent notes

- Prefer live data (metrics, tickets, pages) over generic advice.
- Cite which Google SRE source informed a recommendation when useful.
- Keep outputs tactical: decisions, owners, next actions.

Versions

The database has no other skills named distributed-consensus-reliability.

Database snapshot: 13.11.2026 · package version af4194d98211 (SHA-256 digest of the entire package)

Downloads in the last 30 days: 0 · ranking

Report a problem with this skill: harmful content, incorrect descriptions or an author’s removal request.

Similar skills

• 100distributed-consensus
  Use this skill when the user says 'consensus', 'Raft', 'Paxos', 'leader election', 'split brain', 'quorum', 'etcd', 'ZooKeeper', 'KRaft', 'replicated log', 'linearizable', 'fencing token', 'how many nodes for fault tolerance'. Explains consensus and replication: quorums and…
• 90consensus-basics
  Understand what Raft and Paxos actually provide, quorum arithmetic, and when you need consensus at all. Use when deploying replicated coordination systems or deciding whether a problem truly requires consensus.
• 90distributed-consensus
  Distributed-systems architecture at the protocol level: consensus (Raft, Paxos, BFT), replication and quorums, consistency models, clock synchronization, and the CAP/PACELC trade-offs. Architect-level — how to make state agree and survive failures. USE WHEN: designing…
• 95high-availability-distributed-systems
  Design resilient high-availability distributed systems, active-active multi-region replication, leader election, consensus protocols (Raft), dynamic load balancing, and zero-downtime failover mechanics. Trigger when architecting HA infrastructure, multi-region database…
• 91sre-reliability-engineering
  Use when building reliable and scalable distributed systems.
• 100distributed-systems-interviewer
  A highly theoretical Distinguished Engineer interviewer. Use this agent when you want to test your core distributed systems theory. It probes deeply into the CAP theorem, PACELC, consensus algorithms (Raft/Paxos), clock skew, vector clocks, and how systems manage split-brain…