We have have a primary and secondary ISP connection at the office and configured uplink monitoring to automatically switch between them. In general the primary ISP line has been working well for several months (since Jan 2021) but during the last days we had several short "uplink is down/up" states (for ~1-2minutes) which wos obviously long enough for the UTM monitoring function to notice an outage and engage the backup line.
Our ISP is monitoring the connection from his side and said his system didn't notice any outages. Now they even installed "smaokeping" for our connection to have a more intensive monitoring granularity. However I want't to give them one or two days for getting some statistics before I ask them about the last days outages.
On the other side I'm a bit uncertain if the automatic UTM monitoring could be faulty right now and generate false positives and indicate a "down state" which might be just because some of thir reference servers can not be reached in a certain time (due to internet congestion).
Did someone else experience similar uplink uptages while the automatic monitoring?
Who knows which serves & services are being used for automatic monitoring?
Is there any recommendation which servers should be used for uplink monitoring in case I decide to use my own?
How does the uplink monitoring decide about a down state? When all servers become unreachable or when a certain percantage can't be reached?
In order to get a third opinion I also thought about installing a RasPi/PC with smokeping outside of our Sophos gateway directly on the ISP router/modem (public ip addresses are available) and compare the smokeping statiscics with the detected uplink downs - or does that sound overdone?
This thread was automatically locked due to age.
