Network woes!

[Richie12a]

Hi

New here so please forgive any faux pas:)

My Unraid installation seems to have a problem with its networking. I can ping any of its IP's ( has 2 onboard gb ports and 2 gb ports on an intel card)

What seems to be happening is that Unraid drops its network for a few seconds and then picks up again.

I have attached a snip of the event...1.40 is the IPMI port on the motherboard which appears unaffected, 1.250 is a separate device on the same switch as a control to prove the issue is not with the network hardware/cabling 1.179 and 177 are the connected ports to the Unraid server.

I'm at a loss as to where to go with troubleshooting this...Unraid is latest version....all been cold booted etc....

 

I do see this line in the log but have no idea as to its relevance...

kernel: net_ratelimit: 259 callbacks suppressed

Also I can reproduce the error if I ping a VM on the unraid box!

 

Hoping someone out there might be able to suggest a possible way forward!

 

Many thanks for reading this!

 

 

Model: Custom

M/B: Supermicro - X8DTL

CPU: Intel® Xeon® CPU X5670 @ 2.93GHz

HVM: Enabled

IOMMU: Enabled

Cache: 384 kB, 1536 kB, 12288 kB

Memory: 64 GB (max. installable capacity 192 GB)

Network: eth0: 1000 Mb/s, full duplex, mtu 1500 
 eth1: 1000 Mb/s, full duplex, mtu 1500 
 eth2: not connected
 eth3: not connected

Kernel: Linux 4.18.15-unRAID x86_64

OpenSSL: 1.1.0i

Uptime: 

Edited November 7, 2018 by Richie12a
Added info

[John_M]

You need to post your diagnostics. From the console type diagnostics at the prompt. This will write a zip file in the logs folder on the boot flash device. Shut down and transfer the boot flash to a PC and copy that diagnostics zip. Attach it to your next message here.

[Richie12a]

Apologies!

Here's the diagnostics zip.

hal-diagnostics-20181107-1048.zip

[trurl]

Not sure what you are trying to do. Do you want to bond the 2 ports or what? Maybe try something simpler to start. Just use the default networking and get one port working.

[John_M]

According to your network configuration you have

• a bridge called br0 with one member, eth0, which has a static IP address of 192.168.1.177
• a simple interface, eth1, which has an address allocated by DHCP
• a bridge called br3 with one member, eth3, which has a static address of 192.168.1.179

Now, according to ifconfig you have

• br0 has IP address 192.168.1.177
• br3 has IP address 192.168.1.179
• eth0 is plugged in and working
• eth1 has IP address 192.168.1.22
• eth2 is not connected
• eth3 is not connected

So you have made a mistake in your configuration since the port with address 192.168.1.179 is not plugged in. I suggest you follow @trurl's advice and start afresh.

[Richie12a]

Its quite possible I didnt explain myself properly:)

The issue is that the pings were dropping on a single network connection to the server...I then added a dual nic in an attempt to isolate the problem...179 is the original interface, what seems to be happening is that pings drop on both interfaces to unraid but the IPMI port remains stable which suggests that the issue is within unraid rather than the server hardware. I'm pretty sure this isnt a config issue since the problem pre-dates the second Nic. I'm trying to isolate the issue and I'm reasonably sure that, given the stability of everything not connected directly to Unraid, somewhere in there is the source of the problem..Given my inexperience with Unraid is there a way to preserve my VM's/plugins/configs etc and reload the core software? 

[trurl]

1 hour ago, Richie12a said:

is there a way to preserve my VM's/plugins/configs etc and reload the core software? 

Everything unique to your configuration is stored in the /config folder of flash. You can back that folder up somewhere and prepare that same flash as a new install then restore that /config folder.

 

I am somewhat skeptical that will fix anything though. Unlike other OS such as Windows that you may be familiar with, Unraid effectively does a fresh install each time it boots. The OS is unpacked from the archives on flash and installed into RAM, and it runs in RAM. An update of the OS simply replaces those archives. They are never altered by the OS, so as I said, booting is a fresh install. After boot your configuration (that /config folder) is also read into RAM. And any changes you make to your config in the webUI is written to that /config folder so it can be reapplied next boot.

[Richie12a]

Hmmm..I think I have found the root of the issue...looks like it was a docker network config problem, please consider this closed...thanks for the help!