No Step Closer To A Fix.

[muffintastic]

You forgot this (just in case there are any invisible characters):

nvram set dmz_ip=""
nvram commit

Sorry, copied each single again. My bad.

 

[ColinTaylor]

Can you try making a WAN change now and see if that's fixed it. I don't really expect it to but we've got to check everything.

 

[muffintastic]

Can you try making a WAN change now and see if that's fixed it. I don't really expect it to but we've got to check everything.

I've changed my DNS back to ISP default and had no errors pop up regarding the NAT.

Apr  1 19:36:31 nat: apply nat rules (/tmp/nat_rules_eth0_eth0)

 

[ColinTaylor]

Promising... what does /tmp/nat_rules_eth0_eth0 look like?

 

[muffintastic]

Promising... what does /tmp/nat_rules_eth0_eth0 look like?

I've done two DNS changes, from Quad 9 > ISP back to Quad 9.

This is the results since you told me to input those previous commands .. Even though DMZ was disabled even after a NVRAM reset..

But if I reboot my router and need to change the DNS again etc that error will happen again.

Do you know what has caused this even though DMZ is disabled when factory defaulted or flash new firmware?

 

[ColinTaylor]

Looking good.

But if I reboot my router and need to change the DNS again etc that error will happen again.

Do you know this or are you assuming. Try it now that you've changed that NVRAM variable.

Do you know what has caused this even though DMZ is disabled when factory defaulted or flash new firmware?

Absolutely no idea.

 

[muffintastic]

Looking good.
Do you know this or are you assuming. Try it now that you've changed that NVRAM variable.

Absolutely no idea.

I can only assume since the reboot will probably reset the nvram variable but I will try it when no is using the internet and I will post back to you my result.

 

[RMerlin]

DMZ is disabled. I have no use for it.

But the variable probably contained something rather than being completely empty at it should be. That would explain the invalid rule being generated, as the firmware only checks if that setting is truly empty.

 

[john9527]

But the variable probably contained something rather than being completely empty at it should be. That would explain the invalid rule being generated, as the firmware only checks if that setting is truly empty.

I wonder if there is a glitch in the new nvram routines where instead of empty it's being initialized to a NULL

 

[muffintastic]

Looking good.
Do you know this or are you assuming. Try it now that you've changed that NVRAM variable.

Absolutely no idea.

But the variable probably contained something rather than being completely empty at it should be. That would explain the invalid rule being generated, as the firmware only checks if that setting is truly empty.

I wonder if there is a glitch in the new nvram routines where instead of empty it's being initialized to a NULL

Right guys after a reboot this is the contents of /tmp/nat_rules_eth0_eth0

Now I can't say for sure if this is luck or what but I will give it a day or a least a few hours and if it happens again I will keep you posted, I am not getting my hopes up but i'm not sure what causes this issues even though original DMZ in the WAN section was origionally disabled and still disabled, but I gurantee it will probably error sometime soon. So, unless this can be fixed in a next firmware or looked into deeper because it's a right head scratcher.

I have changed the DNS a few times, disabled upnp and enabled and still no error etc.

 

[ColinTaylor]

Try and remember to check for the error each time you make a change that effects the firewall. That way if it happens again you might be able to isolate what the cause was. Good luck.

 

[muffintastic]

Try and remember to check for the error each time you make a change that effects the firewall. That way if it happens again you might be able to isolate what the cause was. Good luck.

Yes, I'll keep an eye on it in the next couple of days. I wonder if @RMerlin will take a look further into this issue if he gets time or if he's ever came across anything weird like this? I wonder if @john9527 theory is something to look into?

 

[RMerlin]

I wonder if there is a glitch in the new nvram routines where instead of empty it's being initialized to a NULL

A NULL should be the same as "" however when using strcmp().

I want to check if libshared contains any IP validation function, in which case I'll switch the check to that. Otherwise, I might at least make it look for something that's at least 8 chars long (1.1.1.1) to reduce the chances of accidents.

 

[muffintastic]

A NULL should be the same as "" however when using strcmp().

I want to check if libshared contains any IP validation function, in which case I'll switch the check to that. Otherwise, I might at least make it look for something that's at least 8 chars long (1.1.1.1) to reduce the chances of accidents.

Hopefully this is in the next release. Cheers @RMerlin

Update: It happened again when I was adding some ports to the port forwarder added the same tag to /tmp/nat_rules_eth0_eth0 but manage to fix it using Colin's method as a temp fix.