@WalrusWW 

Sorry to hear that. Based on your response my hopes of TP-Link actually solving this issue fell substantially.

We will see what the TP-Link technician can come up with.

  @OneBadGADawg 

I just wish they would add a scheduled reboot feature in the ER605 firmware. I don't run the OC200 or Omada software which does have a reboot feature.

  @caramb 

Replying to myself to provide with some updates.

First, my r605v1 is up and running fine for 29 days now ; so still okay.

Second, on the homelab side, my r605v2 is up and running for 9 days. In a very simple test topology, the issue seems not to occur... (as a reminder, in December, when I first replaced my v1 with my brand new v2, the latter was unable to run for more than a day or two before it apparently hangs and needs a power cylce)

This is quite bad news...

I will follow the plan ; next step will be to generate traffic flowing through the r605v2 and see if it still remain stable. If so, next step will be to setup a more complex configuration similar to the one I use at home.

I continue to see this issue daily ... sometime in the morning the ER605v2 stops working.

From what I see this seems to be DHCP related: When this happens my Macbook can't connect to the Internet and when I do a DHCP renew it doesn't get an IP address. However if i then switch my Macbook to manual IP address it works! It can connect to the Internet again and within the network. 

So it looks like somehow the DHCP server is not responding anymore on the ER605v2, maybe crashed. Unfortunately I can't investigate this further as I have no access to the ER605v2 itself as it's managed via Omada.

Is there any way you can ssh into the router while it's managed by Omada?

  @OneBadGADawg 

only had my er 605 a week, upgraded to latest firmware as a first step as I was waiting for mdns to be supported and I see frequent disconnect and connected messages in the logs, not always in pairs, so a disconnect then reconnect, but maybe 1 disconnect and lots of connected messages. But the odd thing I have is is that my connection doesn't appear to drop. 
 

in fact I have had librenms monitoring it and everything appears fine. One graph that did stand out was the memory consumption. It ramps up and then suddenly drops, so  you get a saw tooth affect, not sure if this could be a symptom that others may not have noticed and could be a factor in the seemingly random times it takes to hang.

but needless to say I'll be watching this thread for developments.

thanks

mkd

  Hi @caramb 

My ER605v1 was put back as my edge router after last weekends firmware fail for the ER605v2. It lasted until Thursday night before starting to disconnect (9 times by the Friday mid morning). I collected the running logs and configuration data from the OC200 and raised a support ticket with TP-Link. 

The process and memory did not seem to rise before the failures, but response time was a clear indicator something was going wrong.

Response time went south until the reboot.

I had an SNMP monitor running for the week as well, but I am looking at lots of counters wondering if the answer is in there somewhere. I think the running logs will be the most helpful for the support team to figure out what is going on. Maybe something sticks out in the SNMP statistics below.

A short update.

The router has hung every day the last days and the pattern I thought I saw earlier with the spike in the CPU utilization was not present the last 2 times. The CPU was low and not spiked as it has been earlier.

But doing diagnostics through the admin portal in the router, the pattern is the same: ping and tracert from LAN is not going through to the Internet but when doing from WAN side everything works. Somehow the router blocks all LAN traffic from getting out.

Nothing new from the conversation with the technician yet, we are just verifying basic configurations and setup. I will try to post interesting information here.

  @Lurk 

Hello,

Thank you for sharing this.

As we don't have command line access to r605v2 with sufficient privileges to properly diagnose the issue, using SNMP to collect metrics is probably the way to go !

I will setup a LibreNMS virtual appliance and share results.

Regarding my homelab, yesterday evening, I setup a sustained 100Mb/s WAN to LAN traffic (so flowing through the r605v2) using iperf3.

The test is running for about 12h now ; still with no issue.

  @caramb 

So, my r605v2 passed the 24h sustained 100Mb/s WAN-to-LAN tcp test.

I setup another one at 200Mb/s.

I the meanwhile, I also setup the libreNMS which is gathering various metrics.