Summary
- Twilio opened incident kntc3sgzlvqf at 22:31:03.878 UTC on 30 July, before the fixed reporting window.
- It warned of SMS delivery delays from Twilio to Africell network subscribers in Angola.
- At 03:13:33.926 UTC on 31 July, Twilio said it had identified the cause but did not disclose it.
- Observed recovery moved the case to monitoring at 04:23:50.427 UTC; further monitoring updates followed at 06:24:20.476, 10:25:34.722 and 18:26:32.087 UTC.
- At the 00:13:55 UTC cutoff on 1 August, there was no resolved timestamp and the case had been open 25 hours, 42 minutes and 51.122 seconds.
- No cause, sender scope, message count, delay distribution, customer count, mitigation or completed recovery was published.
Four recovery notices did not equal resolution
Twilio first reported observed recovery 5 hours, 52 minutes and 46.549 seconds after opening. It then repeated the monitoring assessment three times over the next 14 hours. At cutoff, the latest notice was already 5 hours, 47 minutes and 22.913 seconds old.
The repeated state shows that conditions appeared better but that Twilio had not closed the case. Rewriting monitoring as resolved would erase the operator’s own distinction and use information beyond the fixed reporting boundary.
The cutoff freezes what was knowable
This briefing assesses the incident at 00:13:55 UTC on 1 August. A later status change may eventually appear, but it cannot be imported backward. The correct contemporaneous statement is that resolution had not been published.
Fixed cutoffs are particularly important for long incidents. They let readers distinguish evidence available at publication from later hindsight and make duration calculations reproducible.
Delay remains different from delivery failure
Twilio described delayed SMS delivery toward Africell, not message loss or universal failure. It did not publish a threshold, latency percentile, maximum delay or final delivery outcome.
A delayed message might arrive within or beyond its useful period. Without transaction timestamps, neither the share that remained pending nor the share that ultimately succeeded can be calculated.
Africell defines the destination, not the cause
The notice names Africell network subscribers in Angola. It does not include every Angolan operator and does not describe the Twilio sender numbers, account types or messaging products exposed.
Twilio said the cause was identified at 03:13:33.926 UTC, yet never named it. Africell’s appearance as the destination network is not proof of fault. No intermediary, queue, route or platform component may be assigned responsibility.
The long monitoring phase is itself operational evidence
The first monitoring state began at 04:23:50.427 UTC and remained current at cutoff, a span of 19 hours, 50 minutes and 4.573 seconds. Repeated observations can indicate caution, residual uncertainty or a need for longer stability checks, but the page does not specify which.
The absence of resolution prevents a completed recovery claim. It also leaves open whether some routes, messages or customers still experienced elevated delay.
Expiring messages make latency consequential
SMS authentication, operational alerts and appointment reminders can lose value after a deadline. Applications may retry or switch channel while the original message is still moving, producing duplicates or additional cost.
Twilio did not report those consequences. They explain why a prolonged delay event matters even without proof of permanent message loss. Customer logs must establish any actual workflow impact.
Minor is not a duration-adjusted impact measure
The incident’s minor label stayed unchanged despite its length. That classification supplies neither a volume nor a severity curve. A long incident affecting a narrow route and a shorter, wider incident require different analysis.
No denominator of messages, senders, customers or recipients is public. Duration alone cannot replace it, and the 25-hour clock cannot be converted into customer-hours of impact.
Evidence needed to close the case
A resolved timestamp would establish Twilio’s final state. A post-incident account could then disclose the fault domain, actual impact period, delay distribution, population, recovery action, backlog treatment and preventive work.
At cutoff, the defensible conclusion remains open-ended: Twilio had repeatedly observed recovery for Africell-bound SMS in Angola but continued monitoring. Service improvement was reported; completed resolution and technical explanation were not.
