2

I have recently installed netdata on my Debian 8 server (monitoring software) and afetr installing my Debian 8 server is reporting the below two alerts constantly.

  • 1m ipv4 tcp listen overflows = 37 overflows
  • 1m ipv4 tcp listen drops = 3 drops

I was wondering if anyone could suggest any settings to reduce/stop these errors, maybe some sysctl.conf settings will correct this?

Currenty I have attempted to reduce with the below sysctl.conf settings (with no luck).

net.netfilter.nf_conntrack_max = 524288

net.ipv4.tcp_syncookies = 1

net.ipv4.tcp_max_syn_backlog = 65535

net.ipv4.tcp_max_orphans = 65535

net.ipv4.tcp_synack_retries = 3

net.ipv4.tcp_max_tw_buckets = 1440000

net.ipv4.tcp_tw_recycle = 1

net.ipv4.tcp_tw_reuse = 1

net.core.somaxconn = 65535

net.core.netdev_max_backlog = 65536

net.ipv4.tcp_abort_on_overflow=0

4
  • you need to check the listen backlog in you application man listen and resize net.core.somaxconn
    – c4f4t0r
    Dec 10 '17 at 22:48
  • I have already set the net.core.somaxconn = 65535 to the highest setting?
    – daygle
    Dec 10 '17 at 22:50
  • 1
    what application are you using? because the application needs to set the backlog option when calling listen syscall
    – c4f4t0r
    Dec 10 '17 at 23:15
  • Hmmm I assumed it was Debian and not the application, anyway I'm using github.com/UNOMP/unified-node-open-mining-portal.
    – daygle
    Dec 10 '17 at 23:38

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Browse other questions tagged or ask your own question.