Hello,
I have been experiencing very unreliable FusionHub connections since when upgrading to 8.3.0
I have 3 WAN’s (Starlink, an FWA and a slow ADSL). When monitored one by one, they show random quick disconnections, reason why I mixed them all with a FusionHub instance on AWS.
Outbound policy on my work computer is enforced through FusionHub, but I also tried Priority with FusionHub first, then the other WAN’s.
This has been working seamlessly until fw 8.3.0. I am now showing these behaviors:
- Team calls hang and go frozen or silent on the work computer for half to two minutes; when this happens, the Event log does not even show WAN disconnections; the router is not rebooting, the log looks normal.
- I tweaked the health check of the WAN’s to kind of reflect the actual short disconnections. The logs do not show disconnections when the Team blackouts happen.
- As I keep the same Team call alive in my smartphone, which is connected directly to the WiFi of the FWA router, that WAN is NOT going down, in fact the Team call continues flawlessly on the smartphone: it’s definitely the Peplink
- during at least a couple of these blackouts, when the Team call was frozen, I tried to log into the router web page and it would not load the page; it would then load it when the blackout was passed, as if the router was hanging on something
- I am getting activity in the FusionHub Log, I am trying to paste a screenshot:
The FusionHub WAN is supposed to be “unbreakable”, clearly it is not in my case, any idea of what could be happening here? I am tempted to downgrade to 8.2.0 as it used to work well…
Adding some screenshot of what’s going on, basically unusable while the WAN don’t even disconnect:
In this case, beside the packet loss, the log of the router doesn’t show anything significant:
While this is the log of the FusionHub instance, that also shows nothing between 10:43 and 10:46 when I experienced the disconnections:
Fusinhub settings:
Hello.
Did you check in the status tab, if all links are working with SpeedFusion?
Yes, it looks to me like they are:
I forgot to mention, there is also an LTE but it’s momentarily out of biz while I wait for the new Microtik.
Here is a longish monitor of the FusionHub with PingPlotter, look at all the red lines, all Fusion Hub disconnections:
For comparison, this is the WFA connections alone, directly through the wifi of the Eolo router: only three in a comparative longish time window (not the same time window though):
This is the WFA. The ADSL is actually better, and Starlink, according to the app, has like a disconnection every couple hours, never longer than 2-3 seconds (no obstructions, clear sky).
I would think it’s extremely unlikely for the three connections to fail together, the “unbreakable” FusionHub should be perfect. IT’s definitely something between the Balance and the FusionHub instance…
Hello.
About Starlink… Never seen starlink work 100% of his time. Always some fails during the day, even in a clear sky.
Maybe… We are looking at the wrong direction? Can you monitor the other side?
Can you increase the wan Smoothing at the same link?
I have seen some drops at my network, sometimes, generated when one of links fails, I use this option to minimize the those drops.
I just wanted to give a quick update on this issue. I think I fixed the problem by configuring aggressive wan smoothing, I haven’t experienced disconnections since then, on 8.4.1 now on both Balance 305 and FusionHub.
My take is then that in order to have the “unbreakable” advertised in the marketing material, you need to do wan smoothing.
My Fusion Hub channel is now super stable and averages 60-80 mbit up/down, with the FWA stable at 200/20 and Starlink doing its work in the 200 -250 speed neighborhood. Of course I route through it only the sensitive traffic in order to keep the AWS bill down.
Thank you to peplink technical support and these forums!
1 Like
