WAN Failed DNS Test - Constant Error Since FW 6.3.3 build 3231


#1

Hey everyone -

I have a Peplink Balance 30 that I’ve utilized for a SoHo network for the past few years for two low speed DSL lines out in a rural setup. Everything was working great - never had any issues, outside of an occasional network outage, which wasn’t related to the Peplink.

Within the last 4 months, I upgraded to Firmware 6.3.3 build 3231 - since then, I’ve had constant issues maintaining a steady connection. After running from anywhere in the 20 minute to 4 hour range, all of the sudden the Peplink reports “WAN Failed DNS Test” on both lines and drops all connection.

What’s further odd is that neither DSL model shows any error - in fact, both are still running without issue. After 2-3 minutes of the Peplink reporting the error, the modem jumps in errors and loses connection itself. Thus meaning either the Peplink has some form of ESP, and is detecting a potential error (doubtful), or when the Peplink drops the lines, it sends enough errors to the modem over the WAN port and essentially crashes the modem (much more likely).

Both lines have been tested by the ISP - both at the CLEC and at the NID - there are no reported errors showing. We event went as far as to rerun new internal wiring, and swap all of the connections at the NID. We also replaced each of the modems. By all accounts on the ISP, they’ve gone above and beyond, and my own testing has shown the same results in errors (no CRC or HEC errors until after the Peplink drops the connections).

I also took matters into my own hands and hooked each modem individually up as the uplink for the network - after a day or two, no errors reported and no issues with dropping connection. This also further validates my concerns with the Peplink.

I now see that FW 7 is available - and is my next step in ‘hopefully’ resolving the issue; however, I wanted to see if there was any additional steps that anyone could recommend before I start looking for an alternative replacement.

Thank you in advance!


#2

Can you post what your WAN Health Check settings are? If you are using the ISP’s DNS servers I would recommend trying Google’s instead, 8.8.8.8 and 8.8.4.4

Thanks.


#3

Just a suggestion, why not try changing your health check from DNS to ping and use the default gateway as the ping target? In my opinion and experience, it has been the most reliable method of determining the link health. Depending on how frequent the health check, the DNS server may feel that your health check is being a nuisance and go into auto protect Mode (start dropping requests). PING just seems like a lower cost method of determining link health.

Just my .02. Good luck in finding out the issue.


#4

Negative - we’ve been on Google’s Public DNS since we setup the Peplink - never trust the ISP’s.

Here’s the WAN Health Check settings - pretty much default, outside of the Google Public DNS.

Side note, we upgraded the Firmware to 7.0.0 build 3310 last night to see if it made a difference – it did, kind of – we made it approximately 18 hours before it crashed out again. Same situation as before, it drops WAN, sends out a ton of errors, and crashes the modems.


#5

I’ll definitely give it a try – updated to PING the Gateway for each instead … we’ll see how it goes!