This topic has been marked solved and closed to new posts due to inactivity. We hope you'll join the conversation by posting to an open topic or starting a new one.
How I got Bonding to work with a TP-Link TL-SG3210 switch
Hello,
I thought I would share, since I was puzzled how or why this happened to me and the switch I had. In the end, I got everything to work, and I do NOT think it is a NTGR issue, but if someone else comes across this switch, I have the solution how to make it work. It worked as Layer2, Layer 2+3, and Layer 3+4 options in the end, I left it on 3+4.
The switch I had was a managed TP-Link model TL-SG3210, 10 port gigabit switch with 2 SFP ports. It has a management GUI and a command line interface. The confusing part is that there appears to be 2 ways to make a LAG group with 2+ ports, as there are 2 different tabs to create your LAG. From expereince, all I can tell is method #1 fails, and method #2 works. I repeated this and the same was true.
The first tab is called "Static LAG". This is the way that did not work. I first made the Bond in the network section of the NTGR RNDP6000 (Pro 6), and that was easy. I tried 3+4 right off the bat, and it failed, so I switch to layer 2 only, which also failed. How I got the interface back was to unplug the 2nd ethernet cable, and wait. I am not sure how long to wait, as I went to bed and in the morning it was pingable again, so perhaps 5 minutes? I was dreading doing a factory reset, as I did not have a current NAS backup, though it is said you can reset w/o losing data. I did not want to test that theory it if at all possible. After I could login again, I switched to layer 2 only, same problem, no ping, no access, no NTGR GUI. Once again, I unplugged the 2nd cable, and I could go back into the NTGR GUI.
Here is the 1st SS:
So, I deleted the LAG, and I went into the LACP tab, and checked both ports and changed the drop down to enabled. Seems no different that the other tab, as the other tab you just check 2 boxes and click "Apply". Well, here I checked 2 boxes and did the drop down, but this time I did not have to unplug the 2nd cable, it magically kept working and I never lost ping but for a few seconds while the config reset. After this, I switch the settings in the NTGR sections to Layer 3+4, and tested unplugging either cable while I was pinging, and it worked as intended.
Here is the SS of the correct way to do this:
I am posting this in case another person gets into the same situation. Unplug one of the cables if it stops responding, which should get your access back to make the changes, if needed.
Re: How I got Bonding to work with a TP-Link TL-SG3210 switch
Hi howarddavidp,
This post will surely help others who may encounter the same problem. Tag this thread as resolved by clicking "Accept as Solution" in one of the responses that you received.
The choices were Static LAG or LACP, I guess, but there is little documentation with that switch, so I had to guess.
I agree that the choices are confusing and usually under-documented. You ended up in the right place, and your post will be helpful to others.
Generally the best options with the NAS are LACP (either with a managed switch or an R8500 router), or Adaptive Load Balancing (which works with unmanaged switches). If you don't understand exactly what you are doing, you should simply ignore the other options.
Also, I'd like to point out that the gains for home users aren't very compelling. It's frankly not worth spending a lot of time getting this to work, unless you are in an enterprise environment with lots of users.
Re: How I got Bonding to work with a TP-Link TL-SG3210 switch
Hello,
I am in a quasi-home environment, not a typical home, but also not an office environemnt either.
I have over 20 devices that are constantly connected to my LAN, and at least 15 of those are PC's or MAC's. I have evferything backup with Acronis nightly, between 2AM and 6AM, so at this time I do see high utilization and this LACP connection appears to help with that, judging from how much faster the backup jobs complete vs a single GBe connection.
I am looking into doing this myself, but I have always wished there was an Acronis app for ReadyNAS where I could PXE boot into the device and load Acronis for restores when I have a drive fail (via booting over TFTP to an ISO file stored on my NAS). I know I need to setup a TFTP server (which I have seen available for NTGR), and I also think I would need to make the ReadyNAS the DHCP server, or at least be able to modify the existing DHCP server to pass the "bootfile" parameters in the DHCP info (to clients) to make it work. I have it works on a PC, but not yet with my NAS. My ask would be to have an app that does all of this, but the mods to the DHCP server might be impossible for some who cannot modify this information, but an option wopuld be just use your NAS as the DHCP server. I think people would appreciate this APP if it ever gets made.
I have over 20 devices that are constantly connected to my LAN, and at least 15 of those are PC's or MAC's. I have evferything backup with Acronis nightly, between 2AM and 6AM, so at this time I do see high utilization and this LACP connection appears to help with that, judging from how much faster the backup jobs complete vs a single GBe connection.
15 simultaneous Acronis backups certainly can benefit from bonding.
...I have always wished there was an Acronis app for ReadyNAS where I could PXE boot into the device and load Acronis for restores when I have a drive fail (via booting over TFTP to an ISO file stored on my NAS).
This sounds like a good candidate for the ideas exchange. Maybe post it there.
Re: How I got Bonding to work with a TP-Link TL-SG3210 switch
My backup window (2AM-6Am) was being overrun about 50% of the time, and I get daily emails whether the jobs finished or not.
Since bonding has been enabled, I have yet to exceed the backup window. Not a very technical way to test, but it solves my problem so bonding 100% helps with my environment 🙂
