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

/data-storage

@228d57d
Run it yourself

This skill should be used when the user asks "SQL or NoSQL", "which database", how to design a "data model" or "schema design", picks an "indexing" strategy, needs "sharding" or "partitioning", sets up "replication" (leader-follower / multi-leader), defines a "primary key"/"sort key", asks whether to "denormalize", or weighs "polyglot persistence". Use it whenever a design must decide where records live, how they are keyed and accessed, and how the store scales past one node — even if the user just says "store the data".

Use this Skill: https://skilld.dev/gh/proyecto26/system-design-skills/data-storage

Nothing lands on disk. Nothing to clean up.

Fork this Skill

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

referencesprovidersgcp.md

≈532 tokens on demand. Your agent reads this file only when SKILL.md points to it.

Data Storage — GCP

Service mapping → options

  • Cloud SQL (Postgres/MySQL/SQL Server) — managed relational; backups, read replicas, HA failover. The default managed SQL for single-region scale.
  • Spanner — horizontally-scalable relational: SQL + ACID transactions across shards and regions, with strong (external) consistency. The rare store that scales writes without giving up joins/transactions — at premium cost.
  • Bigtable — wide-column (the original Bigtable); huge write throughput, sorted row keys for range scans. Maps to the wide-column option.
  • Firestore — managed document store with real-time sync; maps to the document option for app/mobile data.
  • Memorystore — managed Redis/Memcached (see caching), not a system of record.

Limits / things that bite (verify against current docs)

  • Bigtable: performance hinges on row-key design — monotonically increasing keys (timestamps, sequential IDs) hot-spot one node ("hotspotting"); use a salted/field-promoted key for even spread. No secondary indexes and no joins — design the row key around your queries.
  • Spanner: avoid monotonic primary keys for the same hot-spotting reason; capacity is provisioned in nodes/processing units and strong cross-region consistency adds commit latency. Powerful but the most expensive option.
  • Cloud SQL: vertical scale and connection limits are bounded by instance size; read replicas are async and can lag.

Provider-specific trade-offs

  • Spanner removes the classic "shard and lose transactions" trade-off but is GCP-only and priced for it — justify with a real cross-shard-transaction need.
  • Bigtable/Firestore lock you into their data and query models; migrating off is real work.

Pitfalls

  • Monotonic row/primary keys in Bigtable or Spanner → hotspotting one node while the cluster looks idle.
  • Reaching for Spanner when a single Cloud SQL node (or replicas) would serve the load — paying for scale you don't need (YAGNI).
  • Expecting Bigtable to do ad-hoc queries or joins — it's query-pattern-first.

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 data-storage skill is an architectural decision-making guide for selecting and designing database systems. It is entirely documentation-based and contains no executable code, network operations, or security risks.

  • 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/data-storage