Auto-negociation messages from kernel on RV320 when forced in 100-full?

[aforster.home]

Hello,

My RV320 has been freezing a lot, and although I have posted in other threads about this continued "Connected/Inactive" issue on WAN1, I didn't see any fix yet. 

This problem has been happening since early versos of the firmware, and still on the current version (1.2.1.13), it is still there.

I noticed a lot of kernel messages and errors about auto-negociation, so I decided to give it a try, and fixed all ports to 100-full, as there is no way to fix them on Gigabit. The weird thing is: kernel messages about auto-negociation error are still there, as follows:

2015-09-06, 00:11:28	Kernel	kernel: nk_bcm53115_sync_speeds_phy_to_mac(): sport[5], speed[1], duplex[2]
2015-09-06, 00:11:28	Kernel	kernel: nk_bcm53115_sync_speeds_phy_to_mac(): sport[5], speed[1], duplex[1]
2015-09-06, 00:11:31	Kernel	kernel: nk_bcm53115_sync_speeds_phy_to_mac(): sport[5], speed[1], duplex[2]
2015-09-06, 00:11:31	Kernel	kernel: nk_bcm53115_sync_speeds_phy_to_mac(): sport[5], speed[1], duplex[1]
2015-09-06, 00:11:33	Kernel	kernel: nk_bcm53115_sync_speeds_phy_to_mac(): sport[5], speed[1], duplex[2]
2015-09-06, 00:11:33	Kernel	kernel: nk_bcm53115_sync_speeds_phy_to_mac(): sport[5], speed[1], duplex[1]
2015-09-06, 00:11:34	Kernel	kernel: nk_bcm53115_sync_speeds_phy_to_mac(): sport[5], speed[1], duplex[2]
2015-09-06, 00:11:34	Kernel	kernel: nk_bcm53115_sync_speeds_phy_to_mac(): sport[5], speed[1], duplex[1]
2015-09-06, 00:11:41	Kernel	kernel: WARNING: no highest common denominator or auto-negotiation not complete
2015-09-06, 00:11:41	Kernel	last message repeated 2 times
2015-09-06, 00:11:41	Kernel	kernel: nk_bcm53115_sync_speeds_phy_to_mac(): sport[5], speed[0], duplex[0]
2015-09-06, 00:11:41	Kernel	kernel: nk_bcm53115_sync_speeds_phy_to_mac(): sport[5], speed[1], duplex[1]
2015-09-06, 00:11:52	Kernel	kernel: nk_bcm53115_sync_speeds_phy_to_mac(): sport[5], speed[1], duplex[2]
2015-09-06, 00:11:52	Kernel	kernel: nk_bcm53115_sync_speeds_phy_to_mac(): sport[5], speed[1], duplex[1]
2015-09-06, 00:11:55	Kernel	kernel: nk_bcm53115_sync_speeds_phy_to_mac(): sport[5], speed[1], duplex[2]
2015-09-06, 00:11:55	Kernel	kernel: nk_bcm53115_sync_speeds_phy_to_mac(): sport[5], speed[1], duplex[1]
2015-09-06, 00:11:56	Kernel	kernel: nk_bcm53115_sync_speeds_phy_to_mac(): sport[5], speed[1], duplex[2]
2015-09-06, 00:11:56	Kernel	kernel: nk_bcm53115_sync_speeds_phy_to_mac(): sport[5], speed[1], duplex[1]
2015-09-06, 00:15:02	Kernel	kernel: wrong ip[0],not_list[0]
2015-09-06, 00:21:10	Kernel	kernel: nk_bcm53115_sync_speeds_phy_to_mac(): sport[5], speed[1], duplex[2]
2015-09-06, 00:21:11	Kernel	kernel: nk_bcm53115_sync_speeds_phy_to_mac(): sport[5], speed[1], duplex[1]
2015-09-06, 00:21:14	Kernel	kernel: nk_bcm53115_sync_speeds_phy_to_mac(): sport[5], speed[1], duplex[2]
2015-09-06, 00:21:15	Kernel	kernel: nk_bcm53115_sync_speeds_phy_to_mac(): sport[5], speed[1], duplex[1]
2015-09-06, 00:21:54	Kernel	kernel: wrong ip[0],not_list[0]

I tried to search for this error message on google, and Google could not find this sync_speeds_phy_to_mac string (believe it or not).

Anyone else has this issue?

[aforster.home]

An update on this thread:

Before changing all ports to 100-full (no auto), I was having freeze issues on the RV320 almost everyday.

It's been 3 days now, since I fixed them to 100-full, with NOT freeze.  My environment is slow as I do a lot of fullHD video streams from my security cameras and 100-full is not enough for my needs, but at least, it is not freezing.

I'd appreciate any feedbacks from anyone having those Connected/Inactive or router freezing issues.

[pahowart]

Having the same issue. I have not tried the 100-full setting yet on my LAN ports as I have the same issue as you with security camera's.

I do have WAN port 1 set to 100-full as it wouldn't connect to my isp with PPOE with the port set to auto.

This is a total BS router. Going to have to send it to the junk heap.

[aforster.home]

I noticed something very weird now.. maybe you could help checking if the same happens on your side.

I had a internet failure now, even though the port is 100-FULL. Looking at the logs I see the last log before the failure is:

2015-09-22, 11:24:49	Kernel	kernel: mac(bc67xxxxf464) not found

I thought "I know this MAC address..." and started looking. And I found this mac address is the the MAC address of my internal vlan port. 

How come the router could not find its own MAC address?  Maybe this is the problem!

One important info: this interface was active during the problem. The only one failed was the internet port, and I just changed to auto negotiation and back to 100-full and it was active again.

Could you check if the same kind of log is found on your side when you have the failure?

Thanks.

[pahowart]

I haven't seen that kernel msg.

I was able to stop the lock ups at least for now buy enabling Energy Efficient Ethernet "EEE" on WAN port 2.

I no longer see the following msg

kernel: nk_bcm53115_sync_speeds_phy_to_mac(): sport[5], speed[1], duplex[2] 

My set up for WAN is now.

Port 1 PPoE DSL - Force 100/Full. (Will not connect with auto)

Port 2 DHCP Cable - EEE enabled Auto.

Now I can focus on finding a proper router. This thing is a piece of junk.

[aforster.home]

The  freezes continue to happen, even with EEE enabled, after upgrade to 1.2.1.14.

[pahowart]

I ended up sending the RV320 to the scrap heap and buying a 6 port china networking box off aliexpress for $200. Loaded pfsense and haven't looked back.

[Francis Sin]

I have found some useful log from my syslog server, it show the log entries just right after crash, I even have the log since "system up" to the time of "crashing".

The excel file attachment is the whole set of syslog since system up to crash.

All the LAN port of the RV320 is unreachable during that time.

Firmware Version is: v1.2.1.14(2015-08-13, 14:21:24), which I believe should be the latest version.

[Francis Sin]

What I should do next.. I want to get the Cisco SMB support team attention to this problem...

looks like something is seriously wrong inside the cisco device OS.

[aforster.home]

I tried to engage them once in the past, but since I dont have a maintenance contract, they seem not to bother much... 

But I would try to open a ticket and see what happens. 

[Francis Sin]

Thanks Aforster, my RV320 also don't have service contract too. >_<

I was try to find a way to create a case about this crash problem half year ago, but I cannot find the place to create a case to them.

I end up gave it up until now >~<, because I got 2 X 1Gbps internet link from 2 different ISP, want to feel how fast can it be.  

[aforster.home]

Can you point me to this 6 port net box? I'm really interested on that as pfsense hw is very expensive. 

Thank you

[Glenn Martin]

Hi.. 

This defects: "CSCuu10991-RV320: WAN port goes inactive intermittently" addresses your issue and a fix has been implemented in F/w rev 1.2.1.14.. 

Thanks,

Glenn

[aforster.home]

Any idea on the freezing problem?

I noticed also that, after moving to 1.2.1.14, my DDNS on WAN1 stops working with "auth failure". After a reboot, it goes back to normal...

[DP007]

We have been having issues with the RV320 and now the RV345. We have 2 ISP connections coming into the router and the LAN switch side will just lock up and stop internet connection the LAN ports. We commissioned the system several years ago and it had been working ok. About 2 months ago it started with the intermittent failures so we bought a new RV320 and replaced it. The same situation occurred where the LAN side of the router will lock up. We can still access the WAN side of the router remotely and get into the router and look at the logs and settings but the LAN side locks up and then it will start working again on its own. We were on with tech support and they indicated they had this issue and that a new firmware was being released soon. They sent us a RV345 because they said that one was not having this reported issue. We installed the RV345 and it is doing the exact same thing. We have fire alarm, security panels and VOIP on the LAN. Nothing that we are using uses any real Bandwidth. The security and fire alarm are so minimal. We have our ports sets to Auto sensing Gigabit. We will switch them to 100. The weird part is that the failure started after so many years and now we can not get rid of the issue even after changing the router twice.