Tutorial on INFINI Pizza

Lesson 1 — First search

Mon, 01 Jan 0001 00:00:00 +0000

Lesson 1 — First search #

Start the node, index documents, run your first queries, and watch them update in real time.

You need a running node — installation covers it in one command. With the node up, index your first document:

curl -X POST -H "Authorization: Bearer $KEY" \
 -H 'Content-Type: application/json' \
 -d '{"message": "hello pizza", "level": "info"}' \
 http://localhost:28000/playground-logs/_doc

There is no refresh call, no flush step — search for it in the same breath:

Lesson 2 — Text search

Mon, 01 Jan 0001 00:00:00 +0000

Lesson 2 — Text search #

How text actually matches: analysis, exact terms, phrases, and the error-tolerant family. Everything below runs live on the logs collection (150 documents with message, level, host, url.path).

Exact terms vs analyzed text #

A term query matches the stored value exactly — no tokenization, no casing help. A match query runs the field’s analyzer first, so casing and punctuation stop mattering:

Compare with a term on the keyword field level — exact, and therefore fast and predictable:

Lesson 3 — Structure and numbers

Mon, 01 Jan 0001 00:00:00 +0000

Lesson 3 — Structure and numbers #

Ranges, sorts, and cursor pagination over the sales collection (120 documents: type, price, qty, amount, merchant, status, date).

The types behind the data #

Numbers are indexed as their real types (float for price, integers for qty), dates as epoch millis. That is what makes the queries below precise — see the field types reference for the full catalog.

Ranges #

A range query takes any combination of gt/gte/lt/lte:

Lesson 4 — Aggregations

Mon, 01 Jan 0001 00:00:00 +0000

Lesson 4 — Aggregations #

From counts to histograms to composite paging — over sales and logs, with known answers you can verify by hand.

Metric aggregations #

One number out: sum, avg, min, max, stats (all of them at once), cardinality (distinct values):

Four distinct product types; stats gives you min/max/sum/avg/count in one round trip.

Lesson 5 — Relations and hybrid search

Mon, 01 Jan 0001 00:00:00 +0000

Lesson 5 — Relations and hybrid search #

The two features that most distinguish Pizza from a plain text index: relations as first-class query targets, and retrieval pipelines you compose explicitly.

Relations #

The articles collection (44 documents) carries a related_to field of type relation — each document declares the articles it points at. A traverse query walks those edges:

Two documents are reachable at depth 0 (the seeds themselves), 17 at depth 1 — the known answer for this dataset. Change max_depth to 2 and watch the circle grow.

Lesson 6 — Operations

Mon, 01 Jan 0001 00:00:00 +0000

Lesson 6 — Operations #

Keeping a node healthy: partial updates in place, compaction, the change feed, and reading cluster state. This lesson is server-side — start a node as in Lesson 1 and run the curls against it.

Partial updates in place #

The document update API takes doc (merge), ops (arithmetic), or both:

curl -X POST -H "Authorization: Bearer $KEY" \
 -H 'Content-Type: application/json' \
 -d '{
 "ops": [
 { "op": "incr", "field": "qty", "value": -1 },
 { "op": "min", "field": "price", "value": 9.9 }
 ]
 }' \
 http://localhost:28000/sales/_update/<doc_id>

Eleven operators — set, unset, incr, mul, min, max, toggle, append, add_to_set, rename, current_date — each a column-level write, not a read-modify-write of the whole document. Fields mapped with "inplace": true apply as O(1) column writes; the full semantics (visibility, arithmetic, mixed updates) are in the partial update reference and in-place columns.