You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
if yes, then the current implementation is incorrect, because, after the current goroutine executed L115, other goroutines could execute L103, which creates a race condition with L116 of the current goroutine, and L103 itself could create race conditions between concurrent goroutines.
After review the code and comments, I don't think StandardEWMA.Tick should be called concurrently, otherwise, it will output wrong results.
for example, after the first tick, even if the race condition is not triggered because we are lucky enough, calling this function from two goroutines at the same time, is just like doing below in a single goroutine:
a.updateRate(a.fetchInstantRate()) // desired updatea.updateRate(a.fetchInstantRate()) // not desired
so we can remove the mutex to make the code simpler and faster.
updateRate and fetchInstanceRate are not public functions, so as long as they're not unexpectedly called by the go-metrics library itself, then they wont race.
Tick() is meant to be called by an external clock every 5 seconds. See meter.go: L226-L251
Do we expect
StandardEWMA.Tick
to be called concurrently?If not, then there's no need to use a mutex and atomic operations on
a.init
, just below code is fine:if yes, then the current implementation is incorrect, because, after the current goroutine executed L115, other goroutines could execute L103, which creates a race condition with L116 of the current goroutine, and L103 itself could create race conditions between concurrent goroutines.
go-metrics/ewma.go
Lines 100 to 120 in cf1acfc
The text was updated successfully, but these errors were encountered: