[rpy]

The government have said the Optus outage was not suspicious. Seems to have been a BGP config change gone wrong.

[jiggawatts]

Is that your best guest or have you seen any insider information?

I haven’t yet seen even the hint of the specific technology involved, let alone a root cause analysis.

[justinclift]

The BGP thing was mentioned on HN as a likely possibility in the main submissions discussing the outage (https://news.ycombinator.com/item?id=38185841, https://news.ycombinator.com/item?id=38185009), and the Cloudflare Radar "BGP Announcements" graph seems to support it.

There's a spike of 940,480 BGP events on Nov 7th at 17:00 UTC, which was 4am Sydney time:

https://radar.cloudflare.com/as4804?dateRange=7d

Note the "7d" (7 days) on the end of the url, so if you're trying that link more than a week after the Optus event you might need to bump that number up.

---

Doesn't seem like captures / snapshots of Cloudflare Radar by The Wayback Machine (web.archive.org), nor archive.today work.

Seems to only capture the html/css/js, but without any data. So the graphs are all empty / grey. :(

[rpy]

They've confirmed it: https://www.abc.net.au/news/2023-11-13/optus-identifies-caus...

[idonotknowwhy]

Would a BGB config cause phones not to connect to the cell towers?

[justinclift]

Maybe the cell towers couldn't load their config's properly or similar?