SBC Intermittent Disconnects

[greatchoice]

We have multiple sites with the raspberry pi as the SBC. We host 3CX on Vultr. I have confirmed there is no DDOS on my VM or at the locations.

Here is the breakdown:

Multiple locations with different ISPs, different phone hardware, different networking equipment disconnect the SBC at the exact same time. It lasts for a few moments and reconnects. However, it happens multiple times back to back.

We have 1 server that hosts 4 separate locations. 2 of the 4 SBC went down at the same time and reconnected with a register OK within 5 seconds but constantly loose registration and gain them back.

We are running 18.1.34 on this location, however, even other locations are running the 18.0.30 version with the same issue.

I cannot figure out any other conciliation at this point. Any advice is appreciated.

 

[SteveITS]

.34 is the latest beta, just for reference. If all 4 SBCs drop at the same time I'd guess it's likely a problem at the server end, perhaps packet loss or a quick drop. We get "SBC up" notification emails for clients and they basically never notice. I'd guess maybe one every week or two? The "up" email is sent on reconnection, but the "down" notification is only sent after 5 minutes or so.

If you're saying it's only two, then it still seems likely to be a problem between those two and the server. Maybe run a "ping -n 7200 x.x.x.x" on both ends to test for packet loss to the other end's router? Often I ping out to multiple places...local router, ISP gateway, something on the Internet, etc., to see where the loss happens.

 

[cobaltit]

So just because there's no DDOS doesn't mean it's not a networking issue. Even with different ISPs if they take the same route to Vultr then it could be a common hop to Vultr. Hard to diagnose without constant connection monitoring between the two