schizoidman@lemmy.zip to Technology@lemmy.worldEnglish · 4 months ago‘There isn’t really another choice:’ Signal chief explains why the encrypted messenger relies on AWSwww.theverge.comexternal-linkmessage-square11linkfedilinkarrow-up1205arrow-down12file-text
arrow-up1203arrow-down1external-link‘There isn’t really another choice:’ Signal chief explains why the encrypted messenger relies on AWSwww.theverge.comschizoidman@lemmy.zip to Technology@lemmy.worldEnglish · 4 months agomessage-square11linkfedilinkfile-text
cross-posted from: https://lemmy.zip/post/51866711 Signal was just one of many services brought down by the AWS outage.
minus-squarejaybone@lemmy.ziplinkfedilinkEnglisharrow-up6arrow-down3·4 months agoWhat reason do they give for only wanting to use those three cloud providers? There are many others.
minus-squareSMillerNL@lemmy.worldlinkfedilinkEnglisharrow-up22·4 months agoScale, they need worldwide coverage. https://mastodon.world/@Mer__edith/115445705126997025
minus-squareCount042@lemmy.mllinkfedilinkEnglisharrow-up2arrow-down2·4 months agoAnd yet a single availability zone in AWS going down caused an outage?
minus-squareSMillerNL@lemmy.worldlinkfedilinkEnglisharrow-up5·4 months agoYes, because scale is not the same as redundancy.
minus-squareJigglySackles@lemmy.worldlinkfedilinkEnglisharrow-up1·4 months agoSignal would need to pay for multi region redundancy and build a programmatic way of swapping regions when their primary region is down.
What reason do they give for only wanting to use those three cloud providers? There are many others.
Scale, they need worldwide coverage.
https://mastodon.world/@Mer__edith/115445705126997025
And yet a single availability zone in AWS going down caused an outage?
Yes, because scale is not the same as redundancy.
Signal would need to pay for multi region redundancy and build a programmatic way of swapping regions when their primary region is down.