@MisterW 

Does the fact that the router doesnt have an arp entry for the printer definitely correspond with not being able to print ?

No, even without ARP entry in the ARP list, I'm able to print / ping. Only after the inactivity of about 10 minutes, I'm not able to ping / print anymore. Then I have to wait some seconds and then it works again. 

So for example with the cron job of ping every 2 minutes, it seems to work.

What protocol is being used to print ?

I'm not sure about that. The POS system is handling this. I'm just using the ping method as that seems easier to debug. Within the router configuration I'm losing packages after +/- 10 min of inactivity. It's the same for these devices: CCV Payment terminal, Epson thermal printer and bixolon label printer.

I now only moved a fedora server and bixolon printer into the Routers network so I can keep testing from home (through VPN tunnel which is all working fine). This way our business can still operate (with the payment terminal, thermal printer and POS connected to the ISP modem)

I'm now testing the pings from the fedora server to the printer. After 10 minutes waiting I get something like this:

If I ping with my macbook (through the VPN tunnel) to the printer, then it is also working perfectly (but this is because it is a connection inside the router probably because of the VPN?). 

And with the cron job enabled (ping every 2 min), it is working perfectly fine (tested for multiple hours). Also with ISP modem/router working fine. 

Things I tried already:

Factory reset: -> same problem

Moving to different subnet .25 -> same problem

Reserve ip -> same problem

Disabling ARP spoofing defense -> same problem

Adding IP-MAC binding -> same problem

Restarting router -> same problem

Changing all lan cables (and tested with cable tester) -> same problem

So I think I almost tried everything except for changing the ARP cache timeout (which I could not find as an option). 

The problem is recreatable with multiple devices. So pinging after +/- 10 min inactivity gives problems with: payment terminal, thermal printer and label printer. 

Tested pinging from macbook (same probleem), from local windows (POS, same problem), server (fedora, same problem). 

Will this ever be fixed in a firmware update? Or is this a hardware problem? We need to decide if we keep this router. 

  @nui_p 

Does the fact that the router doesnt have an arp entry for the printer definitely correspond with not being able to print ?

No, even without ARP entry in the ARP list, I'm able to print / ping. Only after the inactivity of about 10 minutes, I'm not able to ping / print anymore. Then I have to wait some seconds and then it works again. 

Given that , then I think maybe I was barking up the wrong tree, it doesnt appear to be an arp cache problem.

Can you check what the DHCP lease time is set to on the ER706 LAN ?

and maybe also look at the DHCP client list to see if the printers still have a DHCP lease after 10 mins

(I'm assuming the printers are set for DHCP ? )

  @MisterW 

I don't see the printer in the DHCP client list (started happening after I did IP reservation). It is visible in the section "address reservation". 

This is the fedora client in DHCP list:

This is the address reservation section:

DHCP settings: 

  @nui_p

Ah! I see the printer has an address reservation of .15 which is outside the DHCP range of .100 to .199.

Whilst this ought to be OK I'm never sure about that situation. Looking at it from the DHCP server POV how can it allocate an address out of its range...

Could you try deleting one of the printers reservation and giving it a new reservation inside the DHCP range ?

if that works then you could possibly consider extending the DHCP range to cover the addresses you would really like the printers to have or just change the printers IP addresses

  @MisterW 

I thought it was more than OK to do this, because Ip's within DHCP range is handled automatically, and outside the pool you can arrange the stuff (static or reserved)

But I changed the pool to .10 - .199. Afterwards I restarted the router.

Pinged after restart = worked 

Waited 10 minutes = first pings timeout again. 

So no solution :-/

  @nui_p 

I think @Clive_A is right, some detailed diagnosis is required.

Just to review what happens when you ping a device

The typical process is:

1. You enter a command to ping a destination.

2. DNS is used to determine the IP address (if needed).

3. The routing table is consulted to find the next hop towards that destination.

4. ARP is used to find the hardware address of the next hop.

5. The IP packet is sent to the next hop, encapsulated in an Ethernet or WiFi frame.

Now in the case where the destination is in the same network subnet , step 2 is not required and neither is step 3 since the next hop is the device itself. Step 4 is only required if the destination ip is not already in the source device arp cache.

So in the case where you are pinging the printer from a pc, it checks its arp cache to see if it already has the printer MAC address, if not it sends an arp request 'whos got ip address nn' ,having got a response to that it then sends the icmp echo request to that MAC address and waits for a reply.

In theory the router is not involved in this BUT the router is also the network switch. So if the printer and pc are on different network segments then the switch functionality is involved.

I've just checked pinging a device on my network. Initially the arp cache on the pc did not have an entry for the device, pinging updated the pc's arp cache instantly and ping was successfull. Checking the arp cache some 1 1/2 hrs later still showed an entry so it would appear that the pc (windows in this case) maintains the cache well beyond the 10 min you're experiencing.

It might be interesting to check the pc arp cache before and after the ping , and again before trying the ping again after 10 mins plus. That maybe will establish whether its the arp request that's being dropped or the icmp req.