On Fri, Jan 1, 2016 at 9:10 AM, Ben Ripkens <[email protected]> wrote: > Hi, > > I'd like to know how you determine whether the event loop is "healthy" or > "unhealthy", e.g. hitting capacity limits, queue size growing. Do you use > any metrics or statistics? Does libuv expose any metrics for these purposes? > Come to think of it: Queue size sounds interesting to me… > > I have seen people determine an "event loop lag" which can roughly be > summarized as: > > lag = abs(expected time of execution - actual time of execution) > > Would you consider this useful? > > Thanks > > Ben
I don't think there is a need to build in "lag" calculation, libuv users can track that themselves by combining a uv_prepare_t and uv_check_t handle. Libuv currently tracks the number of active handles but not the number of active requests (except through an O(n) operation). I know you are asking in the context of node.js monitoring; quite a few handles in node.js are not exposed to JS land so the current handle count is of questionable use, IMO. You probably want to track HandleWrap instances through process._getActiveHandles(), those correspond more closely to what is visible to JS. -- You received this message because you are subscribed to the Google Groups "libuv" group. To unsubscribe from this group and stop receiving emails from it, send an email to [email protected]. To post to this group, send email to [email protected]. Visit this group at https://groups.google.com/group/libuv. For more options, visit https://groups.google.com/d/optout.
