Affected: API & message delivery, Website & documentation
Resolved ·
One of our servers had a network problem, which caused a brief interruption of the API and website between 17:55 and 17:57 UTC on 16 October, lasting under two minutes. No data was lost. API requests that failed during the interruption should be retried.
Monitoring · · automated
External probes report the service reachable again since 17:56 UTC. Watching before this is called resolved.
Investigating · · automated
External probes report API & message delivery, Website & documentation unreachable since 17:55 UTC. Looking into it.
Recorded by the probes
API & message delivery — 1 min
Website & documentation — 44 s
September 2026
Brief API and website outage during automatic failover
Resolved
—
Affected: API & message delivery, Website & documentation
Resolved ·
Resolved. A server failure made the API and website unavailable for under two minutes (20:45–20:47 UTC). Our systems failed over automatically and service has been running normally since. No data was lost, and Meta redelivered the messages that arrived during the gap.
Monitoring · · automated
External probes report the service reachable again since 20:47 UTC. Watching before this is called resolved.
Investigating · · automated
External probes report API & message delivery, Website & documentation unreachable since 20:46 UTC. Looking into it.
Recorded by the probes
API & message delivery — 50 s
Website & documentation — 50 s
API & message delivery unreachable from external probes
Resolved
—
Affected: API & message delivery
Resolved ·
Resolved. Between 17:48 and 17:50 UTC on 12 September our primary hosting location lost network connectivity. Automatic failover moved the platform to our second location, which has been serving all traffic since; external probes recorded roughly one minute of unreachability. Inbound WhatsApp, Instagram and Messenger messages sent during that window were held by Meta and delivered once we were reachable again; outbound API requests that failed in that window had to be retried. No messages or customer data were lost. Both locations are fully equivalent, so no switch back is required for service continuity. Any future switch between locations is a routine automated operation that completes within seconds, with inbound messages held by Meta in the meantime.
Monitoring · · automated
External probes report the service reachable again since 17:50 UTC. Watching before this is called resolved.
Investigating · · automated
External probes report API & message delivery unreachable since 17:49 UTC. Looking into it.
Recorded by the probes
API & message delivery — 45 s
August 2026
Intermittent availability during a network incident at our hosting provider
Resolved
—
Affected: API & message delivery, Website & documentation
Resolved ·
Retrospective record, added when the Fiwano status page was launched. On 28 August an incident with network equipment at our hosting provider's Vienna data center (provider report: netcup-status.de, 28 August 2026) caused intermittent availability problems for Fiwano over about 48 minutes. Inbound WhatsApp, Instagram and Messenger messages were held by Meta and delivered after connectivity returned; outbound API requests that failed during this period had to be retried.
Resolved ·
Root cause and fix: automatic failover to our second location did not take over. A configuration defect in the failover setup, uncovered during its switch-over that same week, had kept it out of service. The defect was fixed and the database cluster went live across both locations on 29 August. On 31 August a full failover-and-failback drill confirmed automatic recovery in 94 seconds with no data loss, and since 2 September the standby location keeps a pre-built copy of the current release so a failover does not wait on a deploy. A provider incident at one location now means a short automatic switch, not an outage. We also brought forward the launch of this public status page: since 1 September availability is measured by an independent external prober and published continuously, so the record no longer depends on our own reporting.