Hi Toad! First: thx for your work and don't give up the NLM!
First impressions were overall positiv, when I started activating NLM on this node about 138x. My in/out+HTL success rates went up noticeable. Backoffs stayed low. Only thread usage raised a bit +20-30%. Serveral days and versions later thread usage repeatedly climbed beyond my limit after some time of node restart. I always raised the limit; to 500, 750, 1000... Then - when NLM was deployed networkwide with 139x - my thread usage got completly insane >1250 and stayed there! Stopped local activity like FT, sone, filesharing and restartet the node but this doesen't matters much - threadusage was climbing fast again _after some time_. Everytime threadlimit had been reached my node got slow, downloads were staling, backoffs and RejectedOverloads were increasing fast, RequestSender>50%, Clients got disconnected. So maybe this happend to many people, especialy those using the default thread setting. Then most nodes were running on the edge of local threadqueue and continously dropping threads. Whole network traffic could have suffered from "randomly" dropped/not started transfers, which again raised networkwide threadusage itself... I did not look into code - does this make sense ? Maybe just the thread queue/handling need some prio-/optimisation ? In the meantime I reactivated NLM here with 1401: My node seems to run better then ever - and this for 3 day now. Threads <450, HTL >55%, outbound constantly about 75% of limit set, low backoffs & Co. cya Regin
signature.asc
Description: OpenPGP digital signature
_______________________________________________ Devl mailing list Devl@freenetproject.org https://emu.freenetproject.org/cgi-bin/mailman/listinfo/devl