@vGeek79 Just had this happen to me. Have had the switch since January. Had an odd occurrence a few days ago where I would get a constant notification that my OC300 was up then down for a few hours, rebooted everything and it seemed fine. This morning I woke up to one of my APs in isolation mode, I didnt have it set to automatically forward gateway for mesh. I noticed all of my APs had power though. First tried AP reboot, then moving ports, noticed ports 1-4 on switch would all pass power but no link, 5-8 were working as expected. Rebooted switch and lost a couple more ports. Did a full delete and reset though OC300 and now only port 8 will pass traffic. Switch is in a rack with good ventilation so it shouldn't be a heat issue, all fans work however the system light just keeps blinking after reboot. Going to start the replacement process in the morning, hopefully they figure this out. I went this route because of issues with consumer grade products and the amount of devices and data I utilize destroying them, they all lasted longer than this though. 

  @BigD23 

As far as I can tell, the RMA'd device I received is a much more functional revision. However I have not tried power cycling it yet.

  @vGeek79 

Just here to say, I have experienced the exact some problem today too. Restarted the switch and now none of the ports work :( 

  @Utmstgsn Hopefully you get it fixed faster than me. Mine is supposed to be here tomorrow, I paid for next day. It's been over a week now. I went through full reset via console, firmware downgrade, beta firmware provided by support however none of that worked. 

  @BigD23 I called TP-Link to get an RMA ID from them. They didn't waste much time to tell me its defective. I bought it through Amazon so we'll see how smooth that process will be...

I just wanted to update this thread with some more information. I actually had a total of 6 of these switches. When I first posted to this thread, two had failed and they were both running the latest 1.0.6 software. I kept the next two switches at the stock 1.0.2 firmware version, and those worked fine for a few months. After rebooting the switches, they have both now failed.

I gave the last 2 switches away and have moved onto a different brand. Maybe that's extreme, but this is my first TP-Link business-grade product and clearly they're not giving a good impression. I'm curious if the replacement switches are having the same problems.

  @pgookin 

I've looked around on the web and noticed that a lot of people had experienced a power outage or power surge before the switch failed. Can anyone else confirm that they experienced this? Because this was the case for me.

Saw this on Reddit: HELP! TL-SG3210XHP-M2 Managed Switch stopped responding... Cannot ping it, ports don't work. Did full reset through console - still nothing. Only identifies 2 access points on the front, but no internet traffic is sent their way. Any help is highly appreciated! : homelab (reddit.com)

I have seen other posts where people have gone through 3 of them in a year: Third TL-SG3210XHP-M2 in a year : TPLink_Omada (reddit.com)

I ordered a replacement but now I probably should've got a refund. It would be nice if TP-Link could elaborate on what is happening here, it doesn't look like an isolated issue.

  @Utmstgsn No power outage for me. Mine is on a UPS as well. Ports actually started to have some issues which prompted the reboot for me. After that first reboot things got progressivly worse. seemed like everytime I rebooted I would lose more ports. It finally settled on port 8 as the only port that would pass traffic while all the ports would work with POE just fine. We will see if the replacment is any better, I have my doubts. Seems like the unit is not designed very well. 

  @Utmstgsn 

I've gone through 2 in 6 months. The first one was after a power outage, but it was behind a UPS, so there was no surge. It's not like there's an orderly shutdown anyway. The UPS ordered a clean shutdown of the NAS, and then kept the SG3210HXP (and thus three APs and a controller), router, and ONT running until it ran out of battery, at which point it switched off.

The second one I noticed that one of my APs had disconnected. As that's the longest run in the house, I first suspected the cabling, but it checked out OK with a cable tester. I noticed that PoE to that AP was up, but there was no link light on so I rebooted it. A few more ports died. Rebooted again, no copper ports had link. (Power works fine, and the SFPs work fine)

I started a return on both of them, but had to leave on a trip before I could go through the troubleshooting steps that support wanted before issuing an RMA.

It's annoying, because there isn't another option for supporting 2.5G on the EAP660 and 670. This is a critical component of the Omada system and if there's no fix for this problem or sucessor product, I'm going to need to switch systems again. (I'm a recent refugee from UI)