unRAID internet connection slows down

[Fonzarelli]

I'm having a strange problem where my unraid internet connection comes to a grinding halt after a couple of days, it slows down so much that it takes for ever just to download a plugin update.  The strange thing is my docker and vm containers all connect through the same ethernet card and they perform at full speed on the internet with my 150mb internet connection, its only the unraid os that loses almost total internet connectivity.  I have 2 other devices on my wired network that always perform at top speed.  I have tried rebooting unraid and it does not correct the problem, the strange thing is rebooting my internet modem does fix the problem temporary for a couple of days until the speed starts to taper off to almost nothing again. I installed the speedtest plugin and it reports that i have  a Download speed of 0.06 Mbit/s even though all my other devices on the network including dockers ad vm's can utilize the 150mb internet connection.  Any help would be greatly appreciated, Ive been searching for a solution for months. I have attached the diagnostics file.

tower-diagnostics-20180426-2021.zip

[remotevisitor]

My suspicion is that you might have some device on your local network which is using the same IP address as your unRaid system.

 

[John_M]

6 hours ago, remotevisitor said:

My suspicion is that you might have some device on your local network which is using the same IP address as your unRaid system.

 

Well, the server is being allocated its IP address (10.0.1.5/24) by a DHCP server. For this theory to be correct there would need to be some other device that connects occasionally to the network with the same address allocated statically. Before accepting the offer of that address unRAID does a probe to make sure it isn't already in use.

Apr 25 17:11:17 Tower dhcpcd[1520]: br0: soliciting a DHCP lease
Apr 25 17:11:17 Tower dhcpcd[1520]: br0: offered 10.0.1.5 from 10.0.1.1
Apr 25 17:11:17 Tower dhcpcd[1520]: br0: probing address 10.0.1.5/24
Apr 25 17:11:23 Tower dhcpcd[1520]: br0: leased 10.0.1.5 for 604800 seconds
Apr 25 17:11:23 Tower dhcpcd[1520]: br0: adding route to 10.0.1.0/24
Apr 25 17:11:23 Tower dhcpcd[1520]: br0: adding default route via 10.0.1.1

So the other device with the same address would need to connect at some point after this process.

 

@Fonzarelli I notice that the syslog in the diagnostics covers a period of about 26 hours from boot. Did the problem manifest itself during this time? If not you'll need to wait until the problem shows up and then grab new diagnostics, but even then they may not show anything wrong. I see a mention of pihole in your log. Try running without it for a while and see if it is having any bad effect. I have a recollection of it causing some problems for other people recently.

 

[Fonzarelli]

7 hours ago, remotevisitor said:

My suspicion is that you might have some device on your local network which is using the same IP address as your unRaid system.

 

I've used the Fing app on my phone to analyze my home network and the few devices that are on the network have unique ip address. At one point I had a static ip set in unraid and thought that was the problem so now I let my modem assign the ip but still the same issue 

[Fonzarelli]

56 minutes ago, John_M said:

 

Well, the server is being allocated its IP address (10.0.1.5/24) by a DHCP server. For this theory to be correct there would need to be some other device that connects occasionally to the network with the same address allocated statically. Before accepting the offer of that address unRAID does a probe to make sure it isn't already in use.

Apr 25 17:11:17 Tower dhcpcd[1520]: br0: soliciting a DHCP lease
Apr 25 17:11:17 Tower dhcpcd[1520]: br0: offered 10.0.1.5 from 10.0.1.1
Apr 25 17:11:17 Tower dhcpcd[1520]: br0: probing address 10.0.1.5/24
Apr 25 17:11:23 Tower dhcpcd[1520]: br0: leased 10.0.1.5 for 604800 seconds
Apr 25 17:11:23 Tower dhcpcd[1520]: br0: adding route to 10.0.1.0/24
Apr 25 17:11:23 Tower dhcpcd[1520]: br0: adding default route via 10.0.1.1

So the other device with the same address would need to connect at some point after this process.

 

@Fonzarelli I notice that the syslog in the diagnostics covers a period of about 26 hours from boot. Did the problem manifest itself during this time? If not you'll need to wait until the problem shows up and then grab new diagnostics, but even then they may not show anything wrong. I see a mention of pihole in your log. Try running without it for a while and see if it is having any bad effect. I have a recollection of it causing some problems for other people recently.

 

Yes, the last boot 26 hrs ago was to upgrade to the latest unraid version, at the time of upgrade I had to also reboot the modem or it would not download the update. And after 26 hrs unraid has almost no internet connection. Pi hole is a new docker that I set up only a couple weeks ago.. This problem manifested months ago. I have however completely disabled the docker and vm manager for a period of time and the problem still persists. (about 6 months back I did a complete hardware upgrade to the server and just reused the drives, but I can't say for sure if it started right after the upgrade. I do know that I didn't have this problem on my old build).

[John_M]

Wait until it goes wrong again and grab diagnostics before rebooting.

[Fonzarelli]

1 hour ago, John_M said:

Wait until it goes wrong again and grab diagnostics before rebooting.

The diagnostics are from the crippled unraid machine.. The internet slowed down only a  couple of days after the last reboot. I haven't done another reboot because it doesn't seem to fix the problem although rebooting my modem does. 

[John_M]

6 minutes ago, Fonzarelli said:

The internet slowed down only a  couple of days after the last reboot.

 

17 hours ago, Fonzarelli said:

my unraid internet connection comes to a grinding halt after a couple of days

 

The diagnostics don't run for "a couple of days". You said you rebooted both the server and the router yesterday to perform the update and the timestamps confirm that.  The diagnostics need to include the time when you are experiencing the problem and even then they might not show anything useful. I haven't spotted anything in your current diagnostics so wait until it slows down again and we can have another look.

[Fonzarelli]

17 minutes ago, John_M said:

 

 

The diagnostics don't run for "a couple of days". You said you rebooted both the server and the router yesterday to perform the update and the timestamps confirm that.  The diagnostics need to include the time when you are experiencing the problem and even then they might not show anything useful. I haven't spotted anything in your current diagnostics so wait until it slows down again and we can have another look.

Thanks for taking the time to look at this. If it will clean up the logs I will disable docker and vm so you're not going through a bunch of extra code, and then re post the diagnostic file again in a day or two when I experience the slow down.