Re:How key-based cache expiration works
by StavrosK on Sunday February 19, 2012 @07:29AM
(#3609208)
I was thinking the other day, wouldn't it be easyish and optimal to specify/calculate all the data that is used to generate a fragment, store the information in a tree, and invalidate the parents when a node changes.
This would only invalidate data that needs to be invalidated, and it doesn't sound too hard to implement... Does it exist anywhere? What are downsides I haven't thought of?
Re:How key-based cache expiration works
by qwerki on Sunday February 19, 2012 @11:55PM
(#3611349)
not being an expert in this field, one hazard that comes to mind is that a key look up could turn into a tree traversal problem, which defeats the purpose of having a cache in the first place..
This would only invalidate data that needs to be invalidated, and it doesn't sound too hard to implement... Does it exist anywhere? What are downsides I haven't thought of?