Roamze connector status.
Live health of mcp.roamze.com, the Roamze connector for Claude. Checked every minute, with 90 days of history.
mcp.roamze.com
Operational
Last check 54s ago · 306 ms
Last 90 days
UTC days- Operational≥ 99.9%
- Partial outage99–99.9%
- Outage< 99%
- No data
Uptime
Last 24 hours
100.00%
395 checks
Last 7 days
100.00%
395 checks
Last 30 days
100.00%
395 checks
Last 90 days
100.00%
395 checks
Response time
Median (p50)
250 ms
Last 24 hours
95th percentile (p95)
311 ms
Last 24 hours
How this is measured
Every minute, a monitor running inside Roamze’s database requests mcp.roamze.com/mcp/health. A check counts as up only when the connector answers 200 and confirms it can reach its own database — a process that is alive but cannot serve data is recorded as down. Each day’s cell above is the share of that day’s checks that passed.
Checks run from a single vantage point (Roamze’s database in AWS us-east-1), so a network problem between that region and the connector shows here as downtime, and a problem that only affects your own route to it will not. That is the accepted trade-off of a self-hosted monitor over a multi-region one. The monitor runs independently of the connector itself, so it keeps recording through an outage; if it ever stops, this page says monitor stale rather than pretending everything is fine.
Times shown are UTC. Response time is measured end to end from the monitor, including TLS. Learn what the connector does on the connector page, or write to support@roamze.com if something here looks wrong.