My USW Lite 8 PoE and USW Pro XG 8 PoE updated to 7.5.15 and this completely breaks the network. All devices connected to them, including APs experience frequent intermittent connection issues making the whole setup unusable.

• USW Pro XG 8 PoE
• USW Lite 8 PoE
• USW Pro XG 48 PoE (NA)
• EFG 5.0.16 (NA)
• USW Pro XG Aggregation (NA)
• Network 10.2.105

Attachments:

• support-CE61-1775612213455.tgz

Prior version 7.2.123.

Automatic upgrade to 7.4.0 took place March 15th.

Multiple 2.5GbE links to other Unifi hardware (chiefly, a Unifi 6 Enterprise In-Wall) started to flap and display anomalies until downgraded. The links would not remain stable even at 1G speeds. This network has been operational for multiple years.

The problem immediately subsided upon downgrade back to 7.2.123.

Hey folks, when a USW-Pro 24 switch is restarted it shows as offline in the controller but still passes data, and all downstream USW-Pro 24 switches also show as offline and do not come back. The only way to get them back is to delete the "offline" devices, factory reset them, then re-adopt them. But upon doing that, if one switch loses power or needs to be restarted for any reason, that switch and all downstream switches go "offline" again.

However, we can still see upload/download activity on the offline switches in the far right column of the Unifi Devices page, and we see the connected client devices in Client Devices. We also still see the APs as online that are connected to these "offline" switches. We can also see that the switches think they're all good via SSH, all looks right with "info" command.

Because the switches show as offline, we cannot edit their settings, port settings, etc.

Any ideas what could be going on?

Hi, I updated today, and I have the exact same issue you had, did you find a fix for it? I am in a production environment and I cannot factory reset it.

However, we can still see upload/download activity on the offline switches in the far right column of the Unifi Devices page, and we see the connected client devices in Client Devices. We also still see the APs as online that are connected to these "offline" switches. We can also see that the switches think they're all good via SSH, all looks right with "info" command.

Because the switches show as offline, we cannot edit their settings, port settings, etc.

Any ideas what could be going on?

We verified upstream and downstream port configurations, however any device connected to a port assigned to a newly created VLAN would fail to obtain an IP address. Any vlans previously created (from over a month ago) would be able to pass traffic. Only affecting these 2 models every other models in our environment are not affected (USW-PRO-24-POE, USW PRO 48 POE)

Reverting these switches back to 7.2.120 immediately resolved the issue.

Switch models: ECS-48-POE, USW-PRO-XG-48-POE

Previous version: 7.2.120

UniFi Network Application Version: 10.0.162

i would like to downgrade the firmware

On the mobile app it asks for url to enter to downgrade. What/how to find this url?

Many thanks

Peter

I updated my Lite-8-POE from 7.1.26 firmware to this this firmware version. Immediately after the update, two of my three WLANs stopped working. That is, clients couldn't connect to those WLANs and it seemed that they could not get an IP address*. It seemed to be a issue with DHCP.

Some earlier poster stated that they had an issue with this firmware version but it went away by downgrading and updating again. So, I tried that and yes, the problem went away. I do not know what went wrong in the first update but now it seems to work OK.

* I have defined fixed IP addresses for all those clients.

Also, if I enable IGMP Snooping only for a single VLAN, mDNS traffic is negatively impacted across all VLANs. This did not appear to be an issue for my network configuration while on 7.2.26, but was impacted by the migration to 7.2.120 and 7.2.123.

Any help would be appreciated. Happy to participate in troubleshooting if helpful.

ENVIRONMENT:

Gateway:

• UDM-SE

Switches:

• USW Pro XG 48 PoE
• USW Enterprise 48 PoE
• USW Flex XG
• USW Lite 16 PoE
• USW Lite 8 PoE
• USW Flex Mini

Access Points:

• U7 Pro Wall (2)
• U6 Mesh (7)
• IW HD (2)
• AC IW Pro

Bridges

• UDB (2)

I've created a ticket for this. I'll reply back here if this is a known issue.

Attachments:

• network_support_12-01-2026 (1).supp

We reloaded this firmware to test the pro av profiles (multicast) and we did not have any issues, so this must have been re-released with an updated build to correct all the issues that it caused.

Interesting that you don't have issues on this FW version, I am facing a growing latency issue on a network with multiple devices, the latency would creep up to a level when audio cuts, it's the same behaviour like it was discussed in this thread:

Dante latency issues | Ubiquiti Community

I am trying to convince the IT manager to try to downgrade the latest FW version to 7.1.26 as it has solved the Dante issues for some others:

Dante not working on 7.2.123 | Ubiquiti Community

As the problem happens with devices having fairly low network traffic (~5Mbps or so), can it happen that the EEE was introduced to the current release?

The Dante traffic and the control dta for the amplifiers is allocated to a common VLAN, the ProAV Dante profile is used for all ports.

bug 3

unifiOS 4.4.6 (But here, if you roll back to 2 versions, the bug wasn't there.)

network 10.0.0.162 (This bug existed in the previous two versions)

The switches have static IPs. (USW Pro Max 24 PoE)

I remove the client from the VLAN override.

And two switches are stuck in a perpetual getting ready mode.

As soon as I set the switches to DHCP, everything works fine.

bug1 - bug2 - https://community.ui.com/releases/UniFi-Switch-7-2-123/831d8520-c857-44bc-9e6e-2d8ad425e1b5#comment/c56fa34f-86bc-4946-be87-be3b0cfc8725

2 weeks after the update a USW-Industrial connected to a US-XG-16 core switch (using port agg) went offline and stopped passing data. The 2 uplink ports on the XG-16 show online / normal in the controller. Rebooting the XG-16 did not fix the issue. A manual power cycle of the USW-Industrial fixed the problem.

Here’s the weird part:

The exact same thing happened again this morning 2 weeks (to the minute) after the last time. Both times have been on a Sat morning at 3:15AM. No auto updates, nothing obvious happens at the site then. The UniFi logs don’t show anything other than connected devices going offline.

This setup had been stable for years prior to the update.

Any ideas?

UniFi Switches – Improvements in Progress

Hello everyone, and thank you for your ongoing feedback. We want to assure you that we are actively addressing the issues and suggestions you have shared regarding UniFi Switch performance and functionality.

We are currently preparing a set of improvements and fixes, which will be included in the upcoming 7.3 release.

Your feedback is extremely valuable to us. If you are experiencing any issues, please leave a comment or link to an existing topic. Including detailed information and a support file greatly helps us investigate and resolve problems more quickly.

Thank you for your patience and continued support.

Update 12/11/2025. We expect to have the 7.3 Early Access release available soon. This post will be updated with the release link once it is ready.

I am looking forward to this release, I hope we can finally use firmware 7 without issues. Since 7.0.50 we can not update our switches, we have 42 switches, a mix of USW-Ent-48-poe, US-48-POE 500W, USW-PRO-24/48, and USW Pro Aggregation, at the start of version 7.1.26 we have lots of packet loss in all networks and we took 2 weeks of debug to find out that the RSTP in the main USW Pro Aggregations that we have are causing the drop, so we have to leave these on 7.0.50 and the rest on 6.6.61(cause the other versions hangs time to time dropping the traffic). Updating to any major version than 7.0.x caused a lot of false positives RSTP blocks, bad power reports of the switches and the packet drops or even the complete hang of the device, I continuous monitoring the forum and see that these problems still exists on these versions, I still have hope that this work in a near future, I don't like having old firmware but the this is a critical operation problem and as a university I cannot compromise the connectivity to test.

Seems that the problems I mention only exists on big deployments.

It happened again, I plugged the AggPro in, and things went haywire, unplugged it to allow things to fail through my 8 port agg switch, and it seems better now, that the aggpro isn't in the mix. @UI-Team what do you see that I don't. I didn't see anything premature in the logs, no anomalies on ports, etc.

Edit: I moved one of the ports from 25 to 27 to that was connected to my UDM-SE. Things are stable again. It looks like a bad port on the AggPro.

@UI-Team, is there a way to tell if that port I had the UDM connected to is bad by the logs I sent? It would have been port 25 on the Aggregation Pro switch. Unfortunately, that switch is out of stock, or I'd order a replacement, and there's not an equivalent. I am not upgrading to a $2400 switch just because this one is out of stock. When is it coming back into stock? I actually do need one. LOL

Attachments:

• support-DE43-1765217049545.tgz

We have tested on this switches: USW-Pro-24-PoE, USW-Enterprise-8-PoE

@UI-Team