When will these caches be duplicated in memory?

There's a copy in memory, and a copy on disk (For the HTTP cache index and the entire cookie store).  I don't think closing the HTTP cache makes sense, for the main request context.  We do currently have a separate store for media, though (Hopefully going away), and a couple other things sit on top of their own HTTP cache instances, I believe, and closing those seems more reasonable.

Is this in regards to (incognito mode) memory cache? Or just removing the index from memory? I don't think it makes sense to close either cache. In the incognito in-memory cache case, we could shrink it however.

"Either cache" isn't quite right.

There's the main URLRequestContext, and the media URLRequest context, each of which has a disk cache.  Then there's AppCache uses an instance of the disk cache, there's the GPU cache instance, there's a cache + media cache per each isolate app, there's whatever cache is in "Storage\ext\chrome-signin\def\Cache", there's another GPU cache in Storage\ext\chrome-signin\def\GPUCache, there also look to be per-app GPU caches (!).

Right, I assumed this was about the main http cache. How much memory are the others taking up? 

> Right, I assumed this was about the main http cache. How much memory are the others taking up? 

Great question. I am not sure how big a problem these caches are. SimpleCache does show up a lot in traces that I gathered manually on Android. I will try adding instrumentation to see if we can do anything here.

An available bug that is a P3 and hasn't been updated in 180 days. I'm going to archive this. If you disagree with this action, please feel free to reopen and do anything necessary to get someone's attention to the fact that this is still a bug we care about.