Open manju850 opened 2 years ago
I dont have a fix for you, but i can tell you that if your netflow source is handling alot of connections (and thus sending alot of netflow data / IPs / ports TO goNfCollector "stack" , then the stack can see VERY significant disk IO usage, it seems Especially in the beginning of a fresh build. (ive been very impressed with the output goNfcollector provides!)
While i did have some issues during setup/install/spinup - I have NOT seen any crashes nor issues with running goNfCollector for nearly a week- im running it on a Docker standalone setup, on a dedicated Ubuntu20 VM, which runs on a VMware esxi 6.5 host. the disk backing the VM is a high end enterprise NVMe (intel p3700) , its worth noting im only giving the VM ~3gb of ram too.
ive tried adding routers of large buildings i manage , networks for (ie 200+ unit apartments), and at times have seen 20k IOPS , which would overload many non-enterprise setups (and if you have disk IO wait, ram can then be eaten up = OOM / crash more likely). (does not matter what a consumer NVMe disk states as its max IOPS, there are other disk IO considerations beyond a marketing dept's benchmarks IOPS)
sorry i cant help, but perhaps this info will help you look in the correct direction (ie disk io / mem usage)
also what is your OS / version / hardware / ect ? , please.
thanks
Application freeze after around 15-30 mins running. then need to restart . i am not sure its my pc fault or app fault. anyone experienced same issue ??? what is the memory capacity recommended to app. i have 8GB.