How do I scale a DB, wherein I am doing immediate reads rather a lot! (i.e simply after writing/replace I am making an attempt to learn the identical)
I am utilizing AWS RDS and in manufacturing. at present having only one occasion (RW).
though I learn in aws doc that there will likely be some lag within the learn replicas (~19-25ms for me on avg)
I attempted to create few learn replicas and have been utilizing these (I am having node software, so utilizing these with sequleize)
As soon as it is up there, I observed, in my software, there in lots of locations Simply after write/replace I am studying (learn was robotically occurring from reproduction). (instant-read) and on this time I came upon that learn reproduction will not be having the information I am in search of (due to the lag).
I attempted to seek out design patterns on this might not discover any with this sort of use case.
I do know that I can discover out these queries that are doing instant-read and make them run within the grasp as an alternative of a duplicate. however for that, I’ve to jot down/change a lot of locations.
do it? and the way individuals do it? in search of some resolution and or finest observe.