Yeah... I try to pretend that ETS lookup is fast enough to not be relevant. We do config lookups all over the place, so if we're concerned about that I think we need to deal with that as a bigger problem than working around individual lookups. For instance, one option would be to bootstrap new http request pids with a prebuilt config table that has all the options they need to lookup so no concurrent ETS interactions are necessary. Although we would need to do a fair bit of testing to actually determine whether config lookups are a bottleneck.
[ Full content available at: https://github.com/apache/couchdb/pull/1593 ] This message was relayed via gitbox.apache.org for [email protected]
