Index of every page on this site: llms.txt
All skills
proyecto26 avatar

/consistency-coordination

@228d57d
Run it yourself

This skill should be used when the user asks about the "CAP theorem", "PACELC", a "consistency model", "eventual vs strong consistency", "read-your-writes", "causal consistency", "quorum" or "R+W>N", "consensus", "Raft / Paxos", "leader election", "consistent hashing", a "distributed transaction", "2PC", or "saga". Use it whenever a design has multiple copies of data or coordinating nodes and a decision hinges on what a reader is guaranteed to see during replication lag, a network partition, or a node failure — even if the user doesn't say "consistency".

Use this Skill: https://skilld.dev/gh/proyecto26/system-design-skills/consistency-coordination

Nothing lands on disk. Nothing to clean up.

Fork this Skill

Edit a local copy. It keeps the author and licence.

referencesprovidersaws.md

≈529 tokens on demand. Your agent reads this file only when SKILL.md points to it.

Consistency & coordination — AWS

Only the managed pieces that change the generic recipe. For anything not listed, the generic recipe (self-hosted etcd/ZooKeeper, Cassandra, etc.) is the answer.

Service mapping

  • DynamoDB — leaderless, quorum-replicated store with a per-read consistency knob: default eventually consistent reads (cheaper, may lag); pass ConsistentRead=true for a strongly consistent read (latest committed, ~2× the read cost, single-region only). This is the decision-changing detail: consistency is chosen per request, not per table.
  • DynamoDB global tables — multi-region, multi-leader / last-write-wins (AP across regions): every region is writable and converges; cross-region reads are eventually consistent. Strong reads do not span regions.
  • DynamoDB transactions (TransactWriteItems) — ACID across items in one region via an internal 2PC-like protocol; single-region, item-count and size limited.
  • Aurora / RDS — single-leader (writer) + replicas; reads from replicas lag. Multi-AZ failover takes seconds (a write-availability gap).
  • No managed ZooKeeper/etcd as a primitive — run them yourself, or use application-level coordination; MSK ships ZooKeeper/KRaft for Kafka's own use.

Limits / things that bite (verify against current docs)

  • Strong reads in DynamoDB are single-region and don't work on global secondary indexes; global tables are LWW, so concurrent cross-region writes silently lose.
  • DynamoDB partition throughput is per-partition-key — a hot key throttles one partition regardless of table capacity (consistent hashing spreads keys, not a single hot key).
  • Multi-AZ RDS/Aurora failover is seconds, and async replicas can lose the last un-replicated writes.

Pitfalls

  • Assuming DynamoDB reads are strong by default — they are eventual unless you ask.
  • Expecting global-table strong consistency across regions (it's LWW eventual).
  • Treating LWW conflict resolution as safe for counters/accumulators (it isn't — use atomic counters or a CRDT-style model).

Source: SKILL.md on GitHub

No rule matched.

skilld matched fixed text patterns in SKILL.md and file names. Patterns miss obfuscated code.

skilld run checks every file with the same patterns. It asks for approval before it loads a Skill with a behavior marked Needs approval.

No alerts24d3 checks · Risk SAFE
  • Gen Agent Trust Hub24d

    The skill provides instructional content and architectural guidelines for distributed systems (consistency models, coordination, consensus). No security risks, executable code, or malicious patterns were detected.

  • Socket24d

    No alerts

  • Snyk24d

    Risk: LOW · No issues

Signed by skilld at 228d57d. This ties the file your Agent reads to that commit on GitHub. It does not review the instructions.

Last checked against GitHub 3 weeks ago.

Activeupdated 4 weeks ago

README badge

README badge for proyecto26/system-design-skills/consistency-coordination