Strange connectivity issue, when two EAP772v2 are connected to two SG3428XPP-M2

Strange connectivity issue, when two EAP772v2 are connected to two SG3428XPP-M2

Strange connectivity issue, when two EAP772v2 are connected to two SG3428XPP-M2
Strange connectivity issue, when two EAP772v2 are connected to two SG3428XPP-M2
2025-06-13 15:06:56 - last edited 2025-06-16 07:54:55
Model: EAP772  
Hardware Version: V2
Firmware Version: 1.2.2 Build 20250327 Rel. 64822

let me describe our infrastructure:

 

we have three TP-LINK switches:
1x SG2210XMP-M2 v1.0 (1.0.11)
2x SG3428XPP-M2 v1.20 (1.20.11)

and two EAP772v2 (1.2.2)

and SDN Omada (5.15.20.17)

All switches are connected via 10G fiber.

Here is the strange issue now:

One of the EAP772 was version 1, which had an issue with PoE+ and stable 2.5Gbit/s connection. The AP switched automatically from 2.5Gbit/s to 1Gbit/s. Only when a 12V/2.5A power supply was connected, the 2.5Gbit/s connect went well (port settings auto/auto)
TP-Link Support was aware of that issue, and replaced that v1 AP with a v2 version.

Meanwhile another EAP772v2 AP is connected to one SG3428XPP-M2 switch2, setting PoE+ and aut/auto and the connection is stable with 2.5Gbit/s since month.
Today we got the EAP772 v2 replacement, and connected the AP to switch1 SG3428XPP-M2. Same settings as the connected AP on switch2. The new EAP772v2 has now a stable 2.5Gbit/s LAN connection, without any needed power supply.


But now comes the strange issue; suddenly the LAN speed of the EAP772v2, which is connected to switch2, changed the port speed from 2.5Gbit/s to 1Gbit/s themself!!??

So, how can this happen? The second switch AP went well for month and is not directly connected on the same switch where the new EAP772v2 is connected?
Also, we've tried to do a profile override, and set the port speed manually to 2.5Gbit/s and full speed. This drove the LAN AP connection crazy:
HEARTBEAT MISSED..., then PROVISIONING..then DISCONNECTED...PROVISIONING...ADOPTING..CONNECTED...HEARTBEAT MISSED and so. 
Then we setup the port speed back to auto/auto, reboot the AP, the link comes up with 2.5Gbit/s and after a couple of minutes, speed went back to 1Gbit/s.
If we physical disconnect the new EAP772v2 from switch1, the AP speed is good on switch2 with 2.5Gbit/s again.

How can this be? Also we changed the LAN cable (very expensive cat7) already, same issue. It seems to be a PoE+ issue, but how can a connected switch1 AP, affects a switch2 AP, which went well for mounth?

Update: we rebooted both EAP772v2 several times at the same time (totally three times), actual after 42 minutes, both EAP772 are stable with 2.5Gbit/s, just the Rx Dropped Packets increasing a little bit

  0      
  0      
#1
Options
1 Accepted Solution
Resolved:Strange connectivity issue, when two EAP772v2 are connected to two SG3428XPP-M2-Solution
2025-06-16 07:54:42 - last edited 2025-06-16 07:54:55

we did a lot of tests during the weekend, and find a good but strange solution:

in our offices we installed, for nearly one year, thick and expensive CAT6a network cables, those cables went well when one EAP772v2 was connected.
Since we connected the new EAP772v2, to another switch (those switches are connected via 10G fiber), the other EAP772v2 started to reduce the LAN port speed from 2.5Gbit/s to 1Gbit/s, and created thousands of RX errors. 

Our strange solution, we temporary connected that error EAP772v2, from the second switch, to the AP with a cheap!, flat (you can put that cable below doors) CAT7 cable and we were wondering, the connectivity speed was stable with 2.5Gbit/s. What we see is just Rx Dropped Packets (~2004 drops) within 24 hours (those little Rx drops we see on the new EAP772v2 as well, but no Rx errors on both EAPs).

That solution would give us the explanation of a faulty network cable, but do not explain, why it worked couple of month and started problems when a new EAP772v2 is connected to a complete different switch? When the new EAP772v2 would be connected to the same switch, then we could think "OK, it could be a PoE+ issue", but to a different switch?

 FYI, we tested the might be defected network cable, with special cable tester (WireXpert), and couldn't find any PIN problems nor frequency issue.
 

Recommended Solution
  0  
  0  
#4
Options
5 Reply
Re:Strange connectivity issue, when two EAP772v2 are connected to two SG3428XPP-M2
2025-06-14 08:44:52 - last edited 2025-06-14 08:46:00

  @MrEnergy 

Update:

yesterday we restarted the EAP772v2 at the same time, and then the EAP772 were stable, longer as before:

Rebooted:

#2025-06-13 17:31:06,[LLDP]/6/Add a neighbor on port Tw1/0/15.
#2025-06-13 17:30:26,[Link]/5/Tw1/0/15 (GEDU-EAP772-2)  changed state to up.
#2025-06-13 17:30:22,[Link]/5/Tw1/0/15 (GEDU-EAP772-2)  changed state to down.
#2025-06-13 17:30:21,[Link]/5/Tw1/0/15 (GEDU-EAP772-2)  changed state to up.
#2025-06-13 17:30:17,[Link]/5/Tw1/0/15 (GEDU-EAP772-2)  changed state to down.
#2025-06-13 17:30:04,[LLDP]/6/Delete a neighbor from port Tw1/0/15.

this time the EAP772v2 started with the strange issue after 3 hours:

#2025-06-13 21:08:56,[Link]/5/Tw1/0/15 (GEDU-EAP772-2)  changed state to down.
#2025-06-13 21:08:15,[Link]/5/Tw1/0/15 (GEDU-EAP772-2)  changed state to up.
#2025-06-13 21:08:11,[Link]/5/Tw1/0/15 (GEDU-EAP772-2)  changed state to down.
#2025-06-13 20:39:45,[Link]/5/Tw1/0/15 (GEDU-EAP772-2)  changed state to up.
#2025-06-13 20:39:41,[Link]/5/Tw1/0/15 (GEDU-EAP772-2)  changed state to down.

Interesting is, the EAPs, both, are still connected via 2.5Gbit/s, but on the second switch connected EAP772v2 we see these errors now:

 

Rx Packets

21488702484

Rx Bytes

37.46 GB

Rx Dropped Packets

21474838559

Rx Errors

21474836483

Tx Packets

21477091579

Tx Bytes

20.26 GB

Tx Dropped Packets

0

Tx Errors

0

Before we connected the new EAP772v2 to the first switch, that second switch connected EAP77v2 was running fine, without any issues. Any hints?

  0  
  0  
#2
Options
Re:Strange connectivity issue, when two EAP772v2 are connected to two SG3428XPP-M2
2025-06-16 07:38:25

Hi  @MrEnergy 

 

Thanks for posting here.

 

First, I want to confirm:

Has the speed change from 2.5G to 1G issue occurred again after the restart?

 

Regarding the high dropped packets, did you configure VLAN? 

Is there any Internet issue, such as intermittent Internet access or a low-speed issue?

  0  
  0  
#3
Options
Resolved:Strange connectivity issue, when two EAP772v2 are connected to two SG3428XPP-M2-Solution
2025-06-16 07:54:42 - last edited 2025-06-16 07:54:55

we did a lot of tests during the weekend, and find a good but strange solution:

in our offices we installed, for nearly one year, thick and expensive CAT6a network cables, those cables went well when one EAP772v2 was connected.
Since we connected the new EAP772v2, to another switch (those switches are connected via 10G fiber), the other EAP772v2 started to reduce the LAN port speed from 2.5Gbit/s to 1Gbit/s, and created thousands of RX errors. 

Our strange solution, we temporary connected that error EAP772v2, from the second switch, to the AP with a cheap!, flat (you can put that cable below doors) CAT7 cable and we were wondering, the connectivity speed was stable with 2.5Gbit/s. What we see is just Rx Dropped Packets (~2004 drops) within 24 hours (those little Rx drops we see on the new EAP772v2 as well, but no Rx errors on both EAPs).

That solution would give us the explanation of a faulty network cable, but do not explain, why it worked couple of month and started problems when a new EAP772v2 is connected to a complete different switch? When the new EAP772v2 would be connected to the same switch, then we could think "OK, it could be a PoE+ issue", but to a different switch?

 FYI, we tested the might be defected network cable, with special cable tester (WireXpert), and couldn't find any PIN problems nor frequency issue.
 

Recommended Solution
  0  
  0  
#4
Options
Re:Strange connectivity issue, when two EAP772v2 are connected to two SG3428XPP-M2
2025-06-16 08:03:43

  @Vincent-TP 

regarding your questions:

1 Speed: when we rebooted the EAP772v2 it started good with 2.5Gbit/s, then after a couple of minutes (mostly around 15-30 minutes), the ports speed went back to 1Gbit/s and we saw thousands of Rx errors

2. VLAN, yes we use VLANs

3. Internet connectivity: we use double Internet connection with load balancing. Means, to Internet connectivity never lost during our findings with the port speed issue and those RX erros.

 

I guess your questions overlaps with our set final answer, that we changed the network cable from a CAT6a to a CAT7 cable. Actual the port speed seems to be stable (with little Rx drops).

  0  
  0  
#5
Options
Re:Strange connectivity issue, when two EAP772v2 are connected to two SG3428XPP-M2
2025-06-18 13:03:20 - last edited 2025-06-18 13:03:44

Regarding the RX drops, we received an interesting answer from your German TP-Link Support:

Eine geringe Rate an RX verlorenen Paketen, wie zum Beispiel 0,1 %, ist ziemlich üblich und beeinflusst normalerweise den Datenverkehr nicht signifikant. Ich habe einen Screenshot vom EAP aus meinen Tests beigefügt, der ebenfalls eine bemerkenswerte Anzahl an RX verlorenen Paketen zeigt. Trotzdem bleibt der Datenverkehr unbeeinflusst und der Ping ist stabil. Es ist wichtig zu beachten, dass verschiedene EAPs unterschiedliche Chipsets verwenden. Einige Chipsets melden nicht die verlorenen Pakete und zeigen "0" auf dem Controller an. Dies spiegelt nicht die tatsächliche Situation wider; es bedeutet einfach, dass das Chipset die Meldung von verlorenen Paketen nicht unterstützt. Darüber hinaus berechnen Chipsets verschiedener Hersteller "verlorene Pakete" auf verschiedene Weisen. Daher ist ein Vergleich der "verlorenen Pakete"-Daten zwischen verschiedenen EAPs nicht sinnvoll.

  0  
  0  
#6
Options