Hoody SQLite

The Hoody SQLite service is an embedded database runtime co-located with every container. It exposes three complementary surfaces — a SQL engine for relational workloads, a high-throughput key-value store with rich atomic operations, and a set of health endpoints for liveness and observability — all reachable over the container’s SQLite subdomain.

Use these endpoints when you need durable storage inside a container without provisioning an external database. SQL operations back multi-row transactions, the KV store handles counters, queues and snapshots, and the health endpoints feed your dashboards and readiness probes.

Available surfaces

SQL Operations

Run SQL against one or more databases: create databases, execute transactions, and run prepared queries.

Open SQL Operations →

Key-Value Store

Overview of the KV store and listing of keys in a namespace.

Open KV Store →

KV Basic Operations

Get, set, delete, increment, decrement and expiry operations on individual keys.

Open KV Basic Operations →

KV Batch Operations

Batch get, set, and delete operations that round-trip multiple keys in a single request.

Open KV Batch Operations →

KV Atomic Operations

Push, pop, and remove array elements atomically against a single key.

Open KV Atomic Operations →

KV Time-Travel & History

Snapshots, diffs, and rollbacks for KV state recovery.

Open KV Time-Travel & History →

Health

The SQLite service exposes two scrape-friendly endpoints. The first returns a full health snapshot covering service identity, process counters, and the cache subsystem. The second returns only the cache sub-object so dashboards can poll cache pressure cheaply without re-reading process-level data.

`GET /api/v1/sqlite/health`

Liveness/observability endpoint. Returns service identity (status/service/built/started), process-level memory and file-descriptor counters (pid/memory/fds), and hardening snapshots (cache, counters). One scrape covers both liveness and observability signals.

This endpoint takes no parameters.

curl -X GET "https://{projectId}-{containerId}-sqlite-1.{server}.containers.hoody.com/api/v1/sqlite/health" \
  -H "Authorization: Bearer $HOODY_TOKEN"

import { HoodyClient } from 'hoody-sdk';

const client = new HoodyClient({ baseURL: 'https://{projectId}-{containerId}-sqlite-1.{server}.containers.hoody.com', token: process.env.HOODY_TOKEN });

await client.sqlite.health.getHealth();

{
  "status": "ok",
  "service": "sqlite",
  "built": "2026-01-14T09:00:00Z",
  "started": "2026-01-14T10:32:11Z",
  "pid": 42,
  "memory": {
    "rss": 134217728,
    "heapUsed": 41943040,
    "external": 2097152
  },
  "fds": {
    "open": 18,
    "limit": 1024
  },
  "cache": {
    "size": 4096,
    "capacity": 16384,
    "hits": 9821,
    "misses": 174,
    "evictions": 12
  }
}

`GET /api/v1/sqlite/health/cache`

Returns only the cache sub-object from /health. Dedicated endpoint for dashboards that poll cache pressure without re-reading process-level memory and file-descriptor counters.

This endpoint takes no parameters.

curl -X GET "https://{projectId}-{containerId}-sqlite-1.{server}.containers.hoody.com/api/v1/sqlite/health/cache" \
  -H "Authorization: Bearer $HOODY_TOKEN"

import { HoodyClient } from 'hoody-sdk';

const client = new HoodyClient({ baseURL: 'https://{projectId}-{containerId}-sqlite-1.{server}.containers.hoody.com', token: process.env.HOODY_TOKEN });

await client.sqlite.health.getHealthCache();

{
  "size": 4096,
  "capacity": 16384,
  "hits": 9821,
  "misses": 174,
  "evictions": 12,
  "oldestEntryAgeMs": 4210,
  "hotKeys": ["user:67e89abc123def456789abcd", "session:890abcdef12345678901cdef"]
}