Metrics

Per-tenant counters useful for usage-based billing and operational dashboards. Hot writes hit Redis; a daily flush rolls them into tenant_metrics for long-term aggregation.

Configuration

node ace configure @adonisjs-lasagna/saas-tenancy --with=metrics

Writing

import { MetricsService } from '@adonisjs-lasagna/saas-tenancy/services'

const metrics = await app.container.make(MetricsService)

await metrics.increment(tenant.id, 'api.requests')
await metrics.increment(tenant.id, 'api.requests.errors', 5)
await metrics.gauge(tenant.id, 'storage.bytes', usedBytes)

Counters are aggregated per UTC day. Gauges store the latest value.

Flushing

Hot writes accumulate in Redis under tenants/<id>/metrics/<day>/<key>. The flush command rolls them into the database:

# Cron: every night at 01:00 UTC
0 1 * * * node ace tenant:metrics:flush

# Backfill a specific day
node ace tenant:metrics:flush 2026-04-30

The flush uses a SCAN cursor; safe to run against arbitrarily large key sets without the KEYS pattern's gotchas.

Reading

const usage = await metrics.range(tenant.id, 'api.requests', {
  from: '2026-04-01',
  to: '2026-04-30',
})
// [{ day: '2026-04-01', value: 12345 }, …]

Admin REST

GET /admin/multitenancy/tenants/{id}/metrics?key=api.requests&from=2026-04-01&to=2026-04-30

Caveats

• Counter increments do not write to the database; they hit Redis. If Redis is unavailable, increments are lost.
• The flush runs once per day. Until it runs, the database holds no rows for the current day; reads merge Redis + database transparently.
• Don't use this satellite for application-level metrics like request latency. Lasagna ships an OpenTelemetry integration. use the right tool.