DHCP rewrite queue


#1

This is mainly meant for the Sonar Team…
While having large outages (either due to scheduled maintenance or unforeseen problems) the Sonar instance is continuously trying to rewrite the address lists and the DHCP leases to the Mikrotiks, however, due to the outage, the Mikrotiks are no longer reachable. After the outage has been resolved, the number of queued rewrites has increased to a point where our Sonar instance is no longer able to keep up. At this point, no rewrites are happening until I call into to Support to have them “clear” the stale queued rewrites.
We need some mechanism to manually see and “reset” the queued rewrites.
This problem has become more of an issue then the actual tower outage(s).

Any feedback from Sonar?

JB


#2

It shouldn’t make any requests while the devices are unreachable.

If you have some examples or prior tickets where you’ve worked through this, please send them in to support so we can troubleshoot this with you and figure out what’s going on. Just mention this thread and I’ll work with the support team to sort through it.


#3

We have had this issue as long as we have been using Sonar.
My suggestion about a year ago was to make the queue management such that “manually triggered events” like adding/removing IPs to user accounts, get put in the queue at a high priority.
While “auto generated events” like inline device becoming reachable after an outage, get lower priority in the queue.