WAN Fails Test, Connection Up & Down, Peplink 20x

We replaced a Peplink 20 w/ a Peplink 20x in hopes we could stop WAN “disconnects” when it fails a Ping or DNS check. Our ISP (Cox) has replaced 3 modems trying to help us, so we are fairly sure it’s on our side. Well, the switchout did not stop the failures. We have not isolated what causes the failure to know why, so we can correct it. Our big problem has been compounded recently trying to Livestream video when one and often multiple “disconnects” happen causing the streams to fail. It’s been a long time in trying to repair this, so I am reaching out here for some hope we can soon resolve this very troubling issue. Thank you!

Hi. Welcome to the forum!
How many WANs do you have connected to the Balance in total? If its just one then disable the WAN healthcheck.

If its more than one, then when it happens, download a diagnostic file and log a ticket. Engineering can take a closer look to see what the root cause is.

In my experience though - the healthcheck results tend to be accurate and are just highlighting issues with the ISP network that would not otherwise be noticed by normal routers / users…

ISP DNS servers can be a poor target for healthchecks, use google as a test (8.8.8.8, 8.8.4.4)

1 Like

Thanks! We have one WAN. I will disable the health check (we had it set to ping Google’s servers) and try that. Last weekend, the ISP ran a 72-hour monitor of their modem, and they said it was communicating well with no packet loss, however our Peplink B20 reported several times the ping test failed. Since then, we replaced the B20 with the 20x, but it also appears to be doing the same. (One plus so far of the 20x is that we are able to get more bandwidth through. The B20 let 150 Mbps through of 300 Mbps that the ISP provides.) Thank you for the quick response!

I am thinking the problem may be our ISP, but they’ve denied it many times. Any thoughts?

Hello @FlyingPreacher,
Is there a modem or some other network edge device (possibly ISP supplied) between your Balance 20x & the internet connection?

Proving the is a fault with the ISP’s services and getting them to acknowledge it is an art form in itself. The Peplink routers are very useful in capturing the ISP’s reliability of the connection. One of the first things that got us so passionately involved with Peplink was precisely this situation. Proving the ISP’s issue took a good six months of consistent hard work by us plus the client’s IT department. Together we both eventually got the ISP to admit that the ISP supplied edge device was the fault (the brand of the ISP edge device was by the brand that uses a red bridge for its logo). We removed the ISP’s edge device and connected directly to our solution to the client, and the issue was gone.

To prove a fault with the ISP takes time (and lots of it), persistence (you need to have that) and gathering of vast amounts of data.

In Australia we have a National Broadband Network called the NBN. I’m not going to repeat some of what people say the acronym stands for though I do agree with the sentiment of many of those acronyms. Here in Australia, the three major cellular providers are also wholesalers of the government NBN and publicly admit to the NBN issues by their active advertising of supplying routers with failover to the carriers cellular network if the NBN has problems. Peplink|Pepwave routers still outperform the carriers router with a SIM in all tests we’ve done, like comparing chalk and cheese.

So even if you can prove the ISP has a fault, in reality, there may be nothing you or you ISP can do (such as all Australia experiences with the government enforced NBN connections). That is why routers with a 2nd connection to the internet (via cellular for example) are supplied now by many ISPs around the world.

Would you consider adding on a cellular module for your Balance 20x and using InControl2 to help with logging ISP connection issues. You’ll be able to have a greater uptime on your networking service.

You may like to ask your local authorised Peplink Partner for local & personalised assistance.

Your local Certified Peplink Partner can help you with experience on the ISPs in your area and in optimising the ways to use a Cellular backup connection.

Happy to Help,
Marcus :slight_smile:

1 Like

Off the top of my head:

Have you swapped Ethernet cables on your end?

Check the modem from your end. Many respond to 192.168.100.1. The few cable modems that I have checked show counts of both corrected and uncorrected errors. The modem may also have an Event Log.

Peplink routers offer WAN quality monitoring in firmware 8. I am not familiar with it but it could prove useful to you.

If the ISP lets you change out the modem, do so. Get a brand from what they use.

2 Likes

I’ve fixed similar problems with Comcast by using a different DNS server. You are probably getting auto assign DNS through DHCP on the WAN port. Try changing the DNS server to 8.8.8.8 (Google), and use that same address for health check.

2 Likes

Thanks for your help! I am working on getting a wireless modem to provide a backup connection. I am not familiar with logging options with InControl2. Do you have a recommendation as to what to watch for?

We are using Google’s addresses for DNS and health check but am having the same issue.

@Michael234, I will swap out the Ethernet cable and see if that helps. The modem belongs to the ISP (it’s their third one in a year or so), but I do not have access to web into it.

Here’s a picture of the WAN quality when data goes out:


Today, it’s obvious we’ve lost connection several times, but usually when I call our ISP they don’t see it, so they send out a tech, sometimes change out their modem, and say they it should work.

The cable modems that I have dealt with all had web based user interfaces that were accessible from the LAN side. Maybe ask the ISP. Maybe find the manual for the modem.

1 Like

After adding an ATT cellular modem and changing the network cable, we are still having the up and down connection with our ISP. How does one begin the journey of tracking what’s really going on? Our ISP says it might be our problem b/c everything seems to be in levels when they look. Our internal network behind the Peplink 20X seems to be running well. Internet and live-streaming are still experiencing intermittent timeouts. The ATT modem kicks in as a backup, but its 4G data doesn’t always keep data healthy enough for streaming or IP phones. All I want is a healthy data connection. How is that possible?

I have had good luck with using the default gateway as a health check. Depending on the frequency of your health check, your ISP may have some security feature kicking in under the covers. Repeated DNS requests can be perceived as an attack. Ping to the gateway may be a more reliable indicator of the actual link, though this may or may not mean “internet”.

Have you tried disabling the health check entirely? If you don’t notice any issues from a LAN device perspective, you may just be triggering a bunch of false negatives. You can easily run a constant ping to somewhere on the internet from any client device on the LAN and see if you can pick up the drops from there.

I have seen some wonky stuff due to health check failures that cause the Peplink to stop routing traffic to that WAN, but the link looked fine when I did a manual disconnect and reconnect. Best of luck sir.

1 Like

@FlyingPreacher, Recently, I had been having the same issue up until last week or so when I adjusted some of my outbound settings. Since that time, I have not had any issues with the WAN connection dropping out and reconnecting. I have been able to run both beta 2/3 of the most recently released firmware without any issues (WAN drops/device reboots) after making these changes. Whereas, I was only getting a device uptime of 12-26 hours.

Background - I have a 310X, single WAN with Cox (1gig down/40mb up), Verizon SIM in the cell module. I also have a Speedfusion tunnel from this device to a FH Solo Cloud server. This device is also passing through additional Speedfusion tunnels to a secondary Balance device, located just behind it on the local network. My issue was a little worse, as the cellular module activating from standby would force my device to lock up and then restart itself. Not the most optimal situation.

  1. First off you can check your modem from Cox, the ip is usually 192.168.100.1. I own my own device, but the ip address is the same usually. If that doesn’t work, just search google for what the ip address might be.
    1a. If you look at the “Status” page and see a high amount of “Uncorrectables” that could be an issue. From what I read if your line is split too many times, or there is not enough power, or the splitter dies, that will cause issues.
    1b. The “Event Log” page could tell you what the issue might be from there.
    1c. I also could not get through to Cox without spending hours on hold, and I did not have the required equipment to make any changes.

  2. My WAN health check is set to ping the default DNS servers obtained from Cox.

  3. I removed Verizon from standby and just have it active all the time. Due to my reboot issues

  4. I think this might have been what has helped the most and that is adding/adjusting the outbound policies by “Priority,” for each VLAN. Examples below.
    3a. Most of my hardwired devices => Cox - Speedfusion - Verizon
    3b. Wifi VLAN’s => Speedfusion - Cox - Verizon
    3c. IOT VLAN => Strictly Cox
    3d. Cameras => Strictly Cox

1 Like

Hello everyone… I recently replaced my Balance 20 with a Balance 20X and have been having the WAN go in and out frequently. It never happened with the default health checks on the Balance 20. After reading this thread, I tried unchecking use the primary DNS servers and put the 8.8.8.8 and 8.8.4.4 in the healthcheck boxes, and it seemed to be ok but I had WAN disconnects later in the day.

Is anyone else having this issue? Never happened with the 20. Sucks cause every time it happens my WAN IP changes and then I have to change firewall stuff in the cloud…

I’ve been having this problem with my 20x (dropping WAN connection), getting steadily worse. No amount of changing settings has made a difference. In my case, the inbound WAN is fine: when I attach directly to it (not what I want to do, but I can do it), I have NO PROBLEMS with the connection: from all signs this is on the 20x side. Very frustrated.

@david.dorsett,
I was having issues again a few weeks ago, I switched the Health Check Method to “Ping” and left it to ping the default DNS servers picked up from WAN connection. Haven’t had any further issues that I can tell so far.

I have had the best luck pinging the default gateway for each WAN. My logic is that if that device is unreachable, nothing else can be reached. Since it is the first hop, it should always respond to a ping if the link is established.

Your mileage may vary if you do not have a static IP and the gateway changes with new address assignments. If Peplink was to add an option similar to the “Use first two assigned DNS servers as ping hosts” option (only for default gateway) - that would fix it so it works with DHCP scenarios.

I also have a DSL connection that is so horrible that it will sometimes drop PING packets to the gateway. The connection is still up, it is just super saturated while downloading games and stuff. I ended up removing the health check completely on this link. The only down side is that there is a red gripe message on the dashboard. I also don’t have any “priority” or “fall-through” types of outbound policies - so, there isn’t any routing logic based on connection availability. This is a special case where I really don’t care if my gaming PCs and consoles lose connection to the internet. They can wait until the link is re-established. i.e. This strategy only works if you aren’t going to hear any complaints about the internet being down. If anyone is gonna complain, it is best to leave the health checks on and use the priority algorithm.

Hi David, it’s funny you respond to the thread… My 20X started behaving properly with a firmware upgrade and was up for several weeks/months without problems, and the same day you bumped the thread I was getting WAN disconnects again. :frowning:

I’m upset that my decision to build my SMB around Peplink devices vs Ubiquiti was based on a rock solid SOHO MK3 and Balance 20 that was only problematic in its specified bandwidth limitations. Then I plan on rock and rolling with a 20X and it was giving me problems out of the box.