benchmarks

Reproducible benchmarks backing the measurement claims in posts on usamaamjid.com/blog.

If a post here says a number, the script that produced it is in this repo. Run it and check.

mongodb-in-performance.mjs

Backs MongoDB $in Performance: The Array Size Isn't the Problem.

Measures how $in scales as the ObjectId array grows, against 500,000 documents on a real mongod — plan choice, keysExamined, timings, chunking, and the BSON ceiling.

git clone https://github.com/usamaamjid/benchmarks
cd benchmarks
npm install
npm run mongo:in

No MongoDB install needed. mongodb-memory-server downloads a real mongod binary (~80MB, cached after the first run) and tears it down afterwards. A run takes about a minute.

Findings

=== $in on _id (primary index) ===
  IDs        ms    keysExam  docsExam  returned  plan
       10      1.1        11        10        10  FETCH ← IXSCAN
     1000      6.9      1001      1000      1000  FETCH ← IXSCAN
    10000     43.0     10001     10000     10000  FETCH ← IXSCAN
   100000    543.5    100001    100000    100000  FETCH ← IXSCAN

• $in never leaves the index, even at 100,000 values. It expands to one exact-match index bound per value — N point lookups, not a scan.
• keysExamined is exactly n + 1 at every size. No wasted reads.
• Cost tracks documents returned, not array length: a 10-value $in that returns 10,000 docs is slower than a 10,000-value $in returning 10,000 docs.
• Chunking a big $in into batches is slower — same work, more round trips.
• The 16MB BSON ceiling needs ~800,000 ObjectIds. It is not your problem.

Reading the numbers honestly

Timings come from a local mongod with a warm cache and no network, so they are relative, not production figures. Each is the best of three attempts, and even then the same query varies 10–30% between runs. You will get different milliseconds.

What doesn't vary between runs — and what the arguments actually rest on — is the explain() output: plan choice, index bounds, and keysExamined. Those are query planner decisions and they don't care how fast your disk is.

The mongod version is pinned in each script. mongodb-memory-server's default drifts between releases, and a benchmark that silently changes engines isn't reproducible.

License

MIT