Watch report · four sites, six UPSs, five collectors
Two sites need attention. Broadwater’s rack UPS has been on battery for 13 minutes; Shop has been silent for 12.
Duplex is recovering — contact came back at 15:02 and the history it kept while offline is still uploading. Home is steady.
All four sites are reporting. Every UPS is on utility; nothing needs attention.
The last event was Home’s weekly summary, sent at 09:00. Yesterday’s Duplex outage is resolved and kept in the history.
Reporting
3 of 5expected collectors reporting
5 of 6UPSs communicating with their collectors
3incidents open
1,248history records waiting (Duplex)
5 of 5expected collectors reporting
6 of 6UPSs communicating with their collectors
0incidents open
0history records waiting
Incidents
3 open · newest first by severityNone open
last event
No incidents are open. The most recent event on the watch was Home’s weekly summary, sent at 09:00. Yesterday’s outage at Duplex was resolved and remains in its history.
13 min ago
▲Critical · Broadwater
BW Rack UPS on battery
The APC Smart-UPS SMT1500 in the rack has been observed on battery for 13 minutes, on live readings 4 s old. Charge is 71% and the UPS reports 0 V at its input. It estimates 18 minutes of runtime remaining1 — a prediction, not an observation.
Acknowledgment
Not acknowledged
Recovery
Not recovered
Maintenance
Not in maintenance
Notifications
Email + push delivered 15:09:14
12 min ago
◆Warning · Shop
Contact lost with 2 of 2 collectors — cause unknown
Neither shop-bench-win nor shop-esp32-probe has reported since 15:10. We know only that contact stopped; we cannot tell whether power, the network or the machines themselves are responsible.2
Acknowledgment
Not acknowledged
Recovery
Not recovered
Maintenance
Not in maintenance
Notifications
Email delivered 15:15:03
Push failed — retrying
opened15:02recovered
◐Recovering · Duplex
Contact restored; history uploading
Current readings are fresh (3 s). The collector kept a local record while it was out of contact; 1,248 of those records are still on their way. The incident stays open until that history is verified.
Acknowledgment
Acknowledged by jp 13:52
Recovery
Recovered 15:02
Still open
Until backlog verified
Notifications
Delivered
The watch line · last 24 hours
Events at the time they happened
Measured, on utilityMeasured, on batteryEstimate (est.)Delayed record, marked with receipt timeNo recordsRecorded locally, still uploadingUncertain time
Sites · the log
Written from the latest readings
on batterysteady
Broadwater
▲ One UPS on battery · one on utility
● Both UPSs on utility
Collector bw-pi-01 · Raspberry Pi 4 · UPS + Connectivity · contact 4 s ago
bw-desk-ws is asleep. A sleeping workstation is not an outage and opens no incident.
last heardsteady
Shop
○ Contact lost (cause unknown) · 12 min
● Reporting · on utility · fresh
Both designated collectors went quiet at 15:10: shop-bench-win (Windows 11, UPS role) was last heard at 15:10:02, and shop-esp32-probe (ESP32, qualified; connectivity role) at 15:09:58. Nothing since. What follows is the state as it stood then.
Collectors shop-bench-win · Windows 11 · UPS — and shop-esp32-probe · ESP32 (qualified) · Connectivity — both in contact.
Shop Bench UPS · APC Back-UPS Pro 1500 · protects shop-bench-win, shop-cnc-ctrl (no agent), shop-switch
Contact resumed at 15:02 after an interruption that began at 13:41. What the collector sees now reaches us within seconds; what it recorded while cut off is arriving behind it. Between 14:30 and 14:52 there is nothing to arrive — the collector was offline and recorded nothing.
Reporting normally. Yesterday’s outage was resolved and is kept in the Duplex history.
Collector duplex-nuc · Fedora Linux 42 · UPS + Connectivity · contact 3 s ago
Over the last 30 days, observed internet availability was 99.93%, across 98.1% coverage — 13 h 41 m of the period is unknown and not counted either way. Reporting availability for the same 30 days was 99.4%.4