Debugging the GSLB protocol with axdebug and observing communication on port 4149 (gslb protocol port), it appears that 2-3 keepalives every 10-15 seconds is the normal amount. Thousands of keepalives does sound alarming. Is it possible the communication from that member is looping / duplicating in some way? Are the keepalive packets unique?
Thanks for your update. I found that the GSLB keepalives are initiated only by the GSLB master device. Keepalives on member device are low as as they were master for very short duration.
In this case device-1 was master for longest time and has keepalives sent in thousands. while other device, “2nd to master” which became master for short duration has keepalives in hundreds. And the 3rd device which never became the GSLB master only initiated 2 keepalives.
I assume it is alarming if the difference between keepalive sent and received on the same device is high.