Avoid select overhead in hot paths. Just use funcs.
Also, for sort-by-map searches, don't do a describe and pass over all
the results doing location lookups a second time. Remember the
location from the initial matching. Cache it on the search value.
Reduces some sort-by-map searches from 10 seconds to 3 seconds for
me. (still too slow, but good start)
Change-Id: I632954738df9accd802f28364ed11e48ddba0d14
This change allows (for example) to hook a blobserver and an indexer
together in a way similar as they would be in a default Camlistore setup.
This is particularly useful to setup tests that require a realistic
indexer, without having to create a heavier test with e.g. test.World.
This actual goal of this change is described below. I kept these changes
together because the one described below illustrates the need for the
synchandler changes.
pkg/index/stress/: stress test the indexer with more data
The goal of these benchmkarks is to help us compare the various sorted
implementations, so we can pick the most efficient one as a default.
A second, harder, goal would be to add tests that help us estimate the
reliability of the various kv stores, again so we can rule out the buggy
ones, or help fix them.
Change-Id: I8a1fe50398a4ba09c03786502b68b6c2599e5984
Brad Fitzpatrick