Diagnostics

Four read-only commands for "is it up, is my token good, what is it doing".

ping

$ arcli ping
endpoint:   http://localhost:8000 (connection local)
health:     ok (uptime 925.342375ms, latency 4.9ms) storage: hot=ok
auth:       ok as "admin" (id 1, read,write,delete,admin)

ping calls GET /health without a token, then /api/v1/auth/verify with it. Exit 0 means both passed (or the server runs without authentication); exit 1 otherwise. -o json prints the same report as a document even on failure, so a script can distinguish "server down" from "token rejected":

arcli ping -o json | jq -r '[.health.status, .auth.valid] | @tsv'

ping is unauthenticated on the health side, so it never counts as CLI usage in Arc's telemetry (see Privacy).

logs

$ arcli logs --level info --limit 3
┌──────────────────────┬───────┬───────────┬──────────────────────────────┐
│         TIME         │ LEVEL │ COMPONENT │           MESSAGE            │
├──────────────────────┼───────┼───────────┼──────────────────────────────┤
│ 2026-09-08T00:05:37Z │ INFO  │ backup    │ Backup completed             │
│ 2026-09-08T00:05:37Z │ WARN  │ backup    │ Failed to backup config file │
│ 2026-09-08T00:05:37Z │ INFO  │ backup    │ SQLite database backed up    │
└──────────────────────┴───────┴───────────┴──────────────────────────────┘

Reads GET /api/v1/logs (admin token), newest first. Arc keeps the last 10 000 entries in memory per process and returns at most 1000 per call. --level is a minimum (warn returns warn, error and fatal), --since is a window ending now from 1m to 24h (default 1h, rounded up to whole minutes), --limit 1 to 1000 (default 100). Behind a load balancer each call may reach a different node. Messages are what the server captured; one containing a double quote may be cut short by the server's log parser. Table, json or csv (-o csv adds the caller column).

scheduler status

$ arcli scheduler status
continuous queries:  not running (Enterprise license required)
retention:           not running (Enterprise license required)

Whether the continuous-query and retention schedulers are running on this server. On Arc OSS they are not; run retention policies and continuous queries by hand. Any token.

import stats

$ arcli import stats
requests: 1
records:  2
errors:   0

Process-wide counters across every import format since the server started (per node, reset on restart). Any token; documented with arcli import.

All of these take the connection flags.