Re:Slack’s Migration to a Cellular Architecture
by aftbit on Saturday August 26, 2023 @02:29PM
(#37275533)
How can such an architecture function with respect to user data? If the DB instance primary handling your shard is in AZ-1 and AZ-1 gets drained, how can your writes continue to be serviced?
[ Parent
| on HN ]
Re:Slack’s Migration to a Cellular Architecture
by progbits on Saturday August 26, 2023 @02:31PM
(#37275554)
Usually in distributed strongly consistent and durable systems, data is not considered committed until it has been persisted in multiple replicas.
So if one goes down nothing is lost, but capacity and durability is degraded.
Re:Slack’s Migration to a Cellular Architecture
by skybrian on Saturday August 26, 2023 @02:57PM
(#37275807)
That makes sense on its own, but doesn’t it mean that there are lots of network requests happening between silos all the time? It doesn’t seem very siloed.
Or is this some lower-level service that “doesn’t count” somehow?
Re:Slack’s Migration to a Cellular Architecture
by progbits on Saturday August 26, 2023 @03:05PM
(#37275873)
It's siloed that if one is down other's are not affected as long as enough other replicas are healthy to keep the quorum.
You always need cross-AZ traffic, otherwise your data is single homed (which we used to call "your data doesn't exist").