Hello everyone, we’ve been experiencing packet loss on syslog-ng collectors (these are open stack VMs), lately we have increased ingestion and that refelcted packet loss ( especially UDP).

Tuning parameters didn’t help. If anyone here on the list has experienced a similar issue please share your resolution or work around.

We are at 3.9.1, we are planning to update to 3.17.x for better data handling/other bug fixes. As an alternate approach we have increased the VM capacity ( no of VMs).

If there is any that I missed, please point. Any thoughts?

Thank you,

Vijay Amrut.