Hi, I’ve been seeing network sockstat spikes on my primary that now cause errors downstream, replication lag to secondaries and slow performance of the Cluster
I run a 5 member replica set and recently updated to 4.4.23 on Production 17th Aug-
This seems to be occurring during high traffic of 06:00 to around 10:00, I have reduced the load on the primary to read from secondaries, I need assistance on if anyone knows the way to bring these down.
90 Days^
30 Days^
7 Days^
12 hours*