Performance Advisor & collections

Suggested indexes, parsed slow queries and enriched hot namespaces — each deep-linking to a page about one collection.

Performance Advisor & collections in Kubo
Suggested indexes with impact, slow queries parsed into readable pieces, and hot namespaces.

The Performance Advisor is query and index insight without a database driver — it comes from the admin API, so it needs no connection string and no network path to the cluster itself.

Suggested indexes

Each suggestion with its index specification and its impact weight, so you can tell the one that would matter from the three that would not.

Slow queries, made readable

Modern Atlas slow-query log lines are JSON documents, and a phone is the worst place to read one. Kubo parses them into the pieces you actually want:

  • the operation and namespace;
  • duration;
  • the plan summary, with COLLSCAN flagged;
  • the targeting ratio, and examined versus returned counts.

Only the meaningful JSON — the command itself — is rendered, in a collapsible tree, with a Show raw action for the full document when you need it. Alerts and events open their raw JSON the same way.

Hot namespaces, enriched

The raw “hot namespaces” list is just names and counts. Kubo cross-references it with the slow-query and suggested-index signals so each row reads like a finding — “3 slow queries · 2 suggested indexes · collection scan” — and ranks accordingly.

The collection page

Every namespace deep-links to a collection page that pulls everything about one namespace onto one screen:

  • its findings summary;
  • parent-database stats;
  • suggested indexes for it;
  • existing indexes flagged for review — redundant, unused or hidden, from Atlas’ own drop-index suggestions. (The full index list needs a driver, not the admin API; this is the honest subset, and it is the interesting one.)
  • its Atlas Search indexes;
  • its slow queries.

The Collections section browses the same namespaces directly, with a database filter and a name search.

Back to MongoDB Atlas.