Post #3875200
2026-07-17 01:45 UTC
So #Telstra was feeding all their network gear straight from Stratum 1 servers. If they had intermediary Stratum 2 servers in play with constraint facilities, the bad 2006-reset Stratum 1 server would have been ignored, saving the network from complete disaster. #mytake
https://www.theguardian.com/business/2026/jul/17/telstra-missing-software-update-undocumented-design-change-outage
Replies (3)
-
@djb@social.shirow.net 2026-07-17 02:10
@tubsta@social.bsdlab.au if all that’s true, then engineering and operational practices have degraded MASSIVELY
-
@sendai@eigenmagic.net 2026-07-17 01:51
@tubsta@social.bsdlab.au @decryption@aus.social Yeah, soon as I heard that NTP was the root cause, and that they blamed it on one device, that was my first thought. It was such an avoidable failure.
-
@phs@aus.social 2026-07-17 03:35
@tubsta@social.bsdlab.au Regardless of stratum, if clients were configured to be using all 3 servers, shouldn't they have preferred the 2 good ones and ignored the bad 1?