If you're referring to scavenging that kicked in every 2.5 minutes and returned spans which had been unused for 5 minutes, that's been replaced with an asynchronous, continuous mechanism as of Go 1.13 (
https://github.com/golang/go/issues/30333). This policy is generally more aggressive than the old one, so if you're seeing an increase in memory use it's probably coming from something else. Do you know if this is an actual increase in heap objects allocated (a GC trace would tell you (GODEBUG=gctrace=1), also
MemStats.HeapAlloc) or is that roughly the same and the increase is only in OS-reported RSS?
This new policy was also unaffected by the aforementioned bug.
Please file an issue (
https://github.com/golang/go/issues) and I'd be happy to look into it. This isn't the first instance of an increase in memory use we've seen in the issue tracker for Go 1.14, but right now it's hard to tell where it's coming from. It doesn't appear to be the scavenger in any of those other instances, also (AFAICT so far).
Thanks,
Michael