vegeta@lemmy.world to Technology@lemmy.worldEnglish · 2 个月前Musk’s Starlink hit with hours-long outage after rollout of T-Mobile satellite servicewww.cnbc.comexternal-linkmessage-square59fedilinkarrow-up1384arrow-down16
arrow-up1378arrow-down1external-linkMusk’s Starlink hit with hours-long outage after rollout of T-Mobile satellite servicewww.cnbc.comvegeta@lemmy.world to Technology@lemmy.worldEnglish · 2 个月前message-square59fedilink
minus-squarecorroded@lemmy.worldlinkfedilinkEnglisharrow-up25arrow-down3·2 个月前This makes me think that the Starlink system is very poorly designed. I know there are hundreds of satellites, and a large number of base stations. Even if a large chunk of the satellites were taken out and a few base stations failed, shouldn’t the system keep working, just over a different path? This sounds very much not like a hardware failure, but more like somebody fucked up.
minus-squareastrsk@fedia.iolinkfedilinkarrow-up44·2 个月前You’re off ten fold. They have thousands. Around 5000 with a planned 12k after gen 3 has been fully deployed. It’s definitely a “let the intern push to prod” type of scenario by numbers alone.
minus-squarecredo@lemmy.worldlinkfedilinkEnglisharrow-up22·2 个月前They probably dorked up a bgp route or something. It was down globally.
minus-squareHiTekRedNek@lemmy.worldlinkfedilinkEnglisharrow-up3·2 个月前Did you even read the article? It clearly states a core SOFTWARE component. Not hardware.
minus-squareVinstaal0@feddit.nllinkfedilinkEnglisharrow-up3arrow-down1·2 个月前Considering it is designed by an American mega corp, yes it is probably poorly designed because they go for profit maximalisation.
This makes me think that the Starlink system is very poorly designed. I know there are hundreds of satellites, and a large number of base stations.
Even if a large chunk of the satellites were taken out and a few base stations failed, shouldn’t the system keep working, just over a different path?
This sounds very much not like a hardware failure, but more like somebody fucked up.
You’re off ten fold. They have thousands. Around 5000 with a planned 12k after gen 3 has been fully deployed. It’s definitely a “let the intern push to prod” type of scenario by numbers alone.
They probably dorked up a bgp route or something. It was down globally.
That would do it!
Did you even read the article? It clearly states a core SOFTWARE component. Not hardware.
Considering it is designed by an American mega corp, yes it is probably poorly designed because they go for profit maximalisation.