SBC constantly connecting and disconnecting

[HugoWiff]

Sorry if this is the wrong forum, I wasn't sure where I should post about an SBC issue and this seemed like the most appropriate place.


I have a Raspberry Pi 4 Model B (4GB) hosting the SBC running the 3CX Provided Raspberry Pi Image. About once every 5-10 minutes the SBC will disconnected for 0-4 seconds then immediately reconnect, I get 1 monitoring email that it disconnected and another 3-4 that it has 'changed status to UP' before the cycle repeats 5 minutes later.

I have tried swapping the Pi out for a different one and the issue persisted, I took the original malfunctioning Pi back to HQ and the SBC worked fine here.

We have a bit of strange internet setup at the satelite site as we have not been able to get broadband installed for 3-4 months (gotta love NBN here in Aus), so the only connection is via 4G LTE. The 4G modem is connected to the LAN via ethernet and consistent 250/250 bandwidth and 40ms pingtime to google.com. I have not observed any packet loss or latency spikes in any other applications running at the site.

 

[HugoWiff]

Forget to mention in my post that because of aforementioned internet issues, the IP address of the sattelite site is dynamic.

Our primary PBX is cloud hosted by 3CX.

 

[slightlyevolved]

Starting since the around the 22nd of july, I've also been getting hammered with notifications from the SBC "trunk". I never have any emails stating it is down, but I'll get 5-6 emails with the same time saying the SBC Changed Status to UP, then a couple more spread a few minutes after.

I'm not finding anything in the Debian logs on the SBC itself, although, I *could* be missing the correct log file. in our case, both the SBC and our WAN IP are static. (well, technically, dhcp reservation for the SBC)

We are also 3cx hosted, and our SBC is the Debian LInux ISO running under a Proxmox VM host.

 

[HugoWiff]

Starting since the around the 22nd of july, I've also been getting hammered with notifications from the SBC "trunk". I never have any emails stating it is down, but I'll get 5-6 emails with the same time saying the SBC Changed Status to UP, then a couple more spread a few minutes after.

I'm not finding anything in the Debian logs on the SBC itself, although, I *could* be missing the correct log file. in our case, both the SBC and our WAN IP are static. (well, technically, dhcp reservation for the SBC)

We are also 3cx hosted, and our SBC is the Debian LInux ISO running under a Proxmox VM host.

sounds like the same and/or very similar issue

 

[SteveITS]

“Up” triggers whenever it successfully connects. “Down” alerts happen if it’s down a while…want to say it checks every 5 minutes. Repeated “up” alerts just means it has short drops.

Try “ping -n 3600 servername” from the SBC itself and another PC, see if there is loss at the time of the drop.

 

[slightlyevolved]

I'm still investigating, however, I do not think this is the issue as you mention it, because; for example, yesterday I had 6 emails all sent at the same time with the same minute timestamp. That's more like 5 seconds, not 5 minutes. Not a bad idea though, just in this case, doesn't seem to jive with the symptoms.

Could have been some bad flapping of the ethernet switch port, but as it's a VM, some of the other guest should have also been affected. Possible that the SBC is waaay more sensitive.

 

[SteveITS]

yesterday I had 6 emails all sent at the same time with the same minute timestamp. That's more like 5 seconds, not 5 minutes

6 "down" emails? That would be unexpected. I could see 6 "up" emails, sure, that would be 6 one-second drops.

The SBC has a continual connection so will see very short drops, where a web browser will retry a missing packet.

 

[slightlyevolved]

My post says they are UP emails. Not down. Thought maybe I mistyped and just double checked, but no.

It's only sending UP. Like I said, I've been monitoring it. I'm not ruling out some other network issue. Yesterday was the first time it's happened when someone was at work, I absolutely wouldn't rule out our relatively unstable internet service, which is exactly why I went with a hosted solution instead of on-prem.

 

[slightlyevolved]

After an 18hr ping-a-thon, I did get one email stating the SBC Trunk came UP, and checking the log I created, there were NO dropped packets. So I can rule out internal Ethernet issues. Still running WAN tests to see if any drops there match up with the SBC reports.

 

[Wild Edge]

Hi,

We are in Australia and on Starlin satellite with an ADSL failover and have the same issue as you.

Did you get it solved?

Thanks,

Mick

 

[slightlyevolved]

In my case, it kinda fixed itself. I only get notifications like this now when proxmox does a snapshot of the VM, which is fast enough that I never see a down message, but an up for each trunk.

 

[Wild Edge]

In my case, it kinda fixed itself. I only get notifications like this now when proxmox does a snapshot of the VM, which is fast enough that I never see a down message, but an up for each trunk.

Like you, I am not getting the down notifications but am receiving a ridiculously high number of notifications on a daily basis.
We have two remote phones, a Yealink T57W and a Yealink T48U, both set up as remote SBC phones. If we move them back to our home network where our 3CX is hosted, which also has a standalone SBC, everything works with no notifications.Same thing happens with both Yealink units.
Strangely, I am have a remote phone that is not configured as an SBC phone, that is now working remotely, so will see what happens today.

 

[slightlyevolved]

Like you, I am not getting the down notifications but am receiving a ridiculously high number of notifications on a daily basis.
We have two remote phones, a Yealink T57W and a Yealink T48U, both set up as remote SBC phones. If we move them back to our home network where our 3CX is hosted, which also has a standalone SBC, everything works with no notifications.Same thing happens with both Yealink units.
Strangely, I am have a remote phone that is not configured as an SBC phone, that is now working remotely, so will see what happens today.

I found out a while ago that one of the VMs on my Proxmox host was causing high I/O Delay (in the 15-25% range). After resolving that issue, I found that my SBC has been rock solid since with no UP/Down messages. Maybe worth looking at if your SBC device has maybe a failing storage device or similar issue.