TL-SG2008P switch diconnected in omada dashboard that affects WiFi AP2 schedule
Hello,
I need some help to troubleshoot the disconnected TL-SG2008P switch in Omada dashboard.
The TL-SG2008P switch is connected to pfsense firewall through a trunk of several tagged vlans (9, 50,51,53, 56, 59, 61) on port 7. The port 8 is untagged for vlan 9 and it's connected to oc200 controller. The port 6 is a trunk of several tagged vlans (9, 50, 51,59) connected to other manageable switch on other side of the house. Port 1 (POE) is connected to AP1 through a trunk of several tagged vlans (9, 50, 51, 53, 56, 61). Port 2 (POE on schedule in omada controller) connected to AP2 through a trunk of several tagged vlans (9, 50, 51, 53, 56). Port 3 (POE) is powering an AT&T ONT through a poe converter.
The AP1 is working fine (tested on my iPhone by connecting to all 5 SSIDs that using the vlans 50, 51, 53, 56, 61. The ip addresses were correct from each vlan subnets and the internet was working.
The oc200 controller has an ip 10.10.9.15 (from vlan 9) and I could access the Omada dashboard through the ip address and the ip is ping-able from pfsense.
The AP1 ip is 10.10.9.3 (vlan 9), displayed as connected in Omada dashboard and it's ping-able from pfsense.
AP2 is not working because it's connected port 2 is off (port 2 it's on schedule in omada controller) and controller could not see the TL-SG2008P switch to turn on/off the poe port 2. Also the port 2 is off on the switch.
The above setup worked from September 2021 and suddenly disconnected and stay disconnected for 3rd day.
What I did: restarted the controller -same result. Restarted the switch - same result. Disconnected the Ethernet cable from port 7 on the switch and wait 10 seconds and connected back - same result
The switch ip 10.10.9.25 (set in DHCP server on pfsense) is not ping-able from pfsense or from my laptop.
How to troubleshoot in such situation?
I'm thinking to reset the switch (need to do that during the night as it will disrupt the internet for whole house and hope it will get it's assigned 10.10.9.25 ip from pfsense.
Another option is to reset the controller to factory reset, but that is my last resort.
- Copy Link
- Subscribe
- Bookmark
- Report Inappropriate Content
Thanks for all replays,
I have solved the issue by setting al VLANs again on the switch, as I did not remember how I set the first time (a few month before that).
The main point - management VLAN must be untagged, that just means that the device (switch or AP) will get it's ip address from DHCP server as it's connected without any VLAN.
- Copy Link
- Report Inappropriate Content
Dear @ady23 ,
ady23 wrote
The oc200 controller has an ip 10.10.9.15 (from vlan 9) and I could access the Omada dashboard through the ip address and the ip is ping-able from pfsense.
The AP1 ip is 10.10.9.3 (vlan 9), displayed as connected in Omada dashboard and it's ping-able from pfsense.
AP2 is not working because it's connected port 2 is off (port 2 it's on schedule in omada controller) and controller could not see the TL-SG2008P switch to turn on/off the poe port 2. Also the port 2 is off on the switch.The above setup worked from September 2021 and suddenly disconnected and stay disconnected for 3rd day.
To assist you better, I would like to confirm with you that your problem is that the switch is currently showing Disconnected on the controller?
May I know the version of your controller and the firmware version of this switch?
You could try to forget the switch on the controller and see if you can re-add the switch.
Also, are you saying that the PoE Schedule you set is actually in effect, but it doesn't show up in the controller log?
Or does the PoE Schedule not take effect at all?
Best Regards!
- Copy Link
- Report Inappropriate Content
The controller version is 5.0.30 and the switch firmware version is 1.0.2.
I have forget the switch in omada dashboard and nothing changed, so I have reset the switch too. Here what I have found a quite weird switch behavior. As I mentioned before, the switch is connected to pfsense firewall through a trunk port 7 (tagged vlans 9, 50,51,53, 56, 59, 61). For some unknown reason the TL-SG2008P switch is not getting an ip in pfsense or at least I could not find it by switch mac address in the DCHP Lease table.
I have disconnected the switch and connected to a regular subnet (10.10.200.0/24), the switch received right away an ip and I was able to find the ip in the DHCP Leases table on pfsense. I could access the switch web page, also was able to ping it from pfsense. But as soon I connected the switch back to the trunk port to pfsense, I could not find the switch ip address in pfsense. Another step I tried, I have assign manually the all vlans: trunk port 7 (vlans 9,50,51,53,56,59,61), trunk port 1 (vlans 9,50,51,53,56,61), trunk port 2 (vlans 9,50,51,53), trunk port 6 (vlans 9,50,51,59), untagged port 8 for vlan 9 on the switch web page. Again connected the switch to pfsense trunk, AP1, AP2 and it didn't work, the switch still couldn't be find in pfsense DHCP Lease table.
I then tried another 5 port managed switch (Netgear GS105Ev2), I set manually the trunk tagged port 1 (vlans 9,50,51,53,56,59,61) and untagged port 2 for vlan 9, untagged port 3 for vlan 9. I then connected port 1 to the same pfsense trunk (used same exactly Ethernet cable that was connected to TL-SG2008P port 7) and my laptop connected to port 3 received an ip address from vlan 9. OK, that means the pfsense trunk was working fine for vlan 9.
I then connected the oc200 controller to port 3, and the port 2 to port 8 on TL-SG2008P switch (I reset the TL-SG2008P switch again). In this setup the TL-SG2008P received the ip address from pfsense in the vlan 9 subnet (the one assign by pfsense DHCP static setup for switch mac address) and it was discoverable on DHCP Lease table and it's ping-able. The TL-SG2008P switch appeared in Omada dashboard and was adopted. I have made one more setup change on the Netgear GS105Ev2 switch, I have set the port 2 to be tagged trunk for vlans 9,50,51,53,56,59,61. Then in Omada dashboard I have set the TL-SG2008P ports to use the correct profiles: port 1 to use AP1 profile (vlans 9, 50, 51, 53, 56, 61), port 2 to use AP2 profile (vlans 9, 50, 51, 53, 56), port 7 to use the pfsense trunk (vlans 9,50,51,53,56,59,61), port 6 to use other management switch (vlans 9, 50, 51, 59), untagged port 8 to use vlan 9 and untagged port 4 to use vlan 9 (if I will like to connect oc200 to poe port).
Connected the AP1 and AP2 to TL-SG2008P switch on port 1 and 2 and the WiFi was working again, as it should be. Had to add back the poe schedule and it worked as expected.
Based on the above, why the switch is not discoverable in pfsense when connected to pfsense trunk directly, and is working with similar trunk from another managed switch?
Is it other way to assign an ip to TL-SG2008P managed multi vlan switch?
- Copy Link
- Report Inappropriate Content
I believe what you meant is that switch is discovered when it is connected to an untagged ( vlan 1) port and when you move the switch to a trunk uplink that does not have vlan 1 / untagged, then switch is not discovered as it is not able to obtain a DHCP IP.
Yes that is the default behaviour. There is option to set management vlan ID, once switch is discovered on vlan 1 and then switch will save that and then it will start sending a tagged DHCP request off of that management VLAN ID. Then you should be able to move the switch to the trunk port that has tagged vlans only. Just be very careful else you will need to factory reset the switch again and joined then via untagged VLAN 1 port.
- Copy Link
- Report Inappropriate Content
I think you are right. The problem is how the switch and AP are getting the ip addresses and how they use the MANAGEMENT VLAN.
I have set the MANAGEMENT VLAN (vlan 9 in my case) on my TL-SG2008P switch (yes, I have to reset the switch a few times, and not afraid any more) and I have in the config>>services >> MANAGEMENT VLAN: MANAGEMENT_9 (where "MANAGEMENT_9" is my created management vlan , and also on config >> VLAN Interface - MANAGEMENT_9 is enabled.
Even with that it's still not working when the MANAGEMENT VLAN is part of a trunk with other vlans. I have found that It's working if it's a separated on a different port and untagged (I have tried many times and each time as soon I pass the management vlan together with other tagged vlans in a trunk, the switch was disconnecting in Omada dashboard and connected as soon I add the Management VLAN separately untagged (didn't try with separated and tagged, will need to try that).
I have used this instructions to set the management vlan from here: https://www.tp-link.com/us/support/faq/2814/ (topology 2 as I have pfsense firewall). It's not very clear how the switch is connected to a router and how the devices are getting their ip addresses (only how to create and set the MANAGEMENT VLAN and how the controller is connected to a switch and also it's mentioned an Access Point too). Not sure why it didn't work in my case.
I have also similar issues with my 2 EAP660HD (ver. 1.1.0) WiFi AP recently, both are set to have static ip (part of MANAGEMENT_9 vlan subnet), and both are connected to TL-SG2008P switch poe ports. To adapt them I used the MANAGEMENT_9 vlan profile only and after all setup I connect them through a trunk of several vlans for different SSIDs ( I have created a profile that includes multiple vlans (SSID + MANAGEMENT_9). I had also tried to setup the Management VLAN: MANAGEMENT_9 on the AP explicitly in omada (per the instructions mentioned above) but that didn't worked either, as soon I set the Management Vlan to my MANAGEMENT_9 vlan, the Access Point disconnects and is staying in "Adopting" stage for a while in the dashboard and never connected again (need to reset the AP). But that also happens if I don't set the MANAGEMENT VLAN, as I have found later, I see the APs in "Adopting" stage and their ip changed in the dashboard (the changed ip is from one of the vlans and not the static one that was set that is part of the subnet of my vlan 9). The Wifi looks like it's working on different SSIDs, but if the ip is incorrect, the dashboard is not displaying the info from the Access Points.
Very frustrating to deal with different vlans in omada. I have used 4 vlans for 4 SSID on my tomato shibby router for more than 5-6 years and later when moved to pfsense, used 2 tomato router as Access Points and that setup was just working.
- Copy Link
- Report Inappropriate Content
@ady23 Did you find the problem in the mean time?
I have the same switch TL-SG2008P ( and a TL-SG2008 and TL-SG2218) and using a OC200 controller.
I was playing with pfSense yesterday and i had the same problem. I saw that all the 3 switches showed as status disconnect on the OC200 dashboard.
But they still work.
I use Vlan1, 10, 124, 125 and 126.
I removed pfSense but the switches did not connect anymore. I did a power cylce for each switch and they came back.
But last week i was also playing with pfSense and did not had a problem. But the VLAN's did not work because i forgot to set the VLAN port on pfSense to 4095 (all).
Next it tested with 1 vlan. VLAN10. This was not configured in the OC200 controller. So i added the VLAN but not as a interface but as VLAN.
This works and i did get a IP from the DHCP server on pfSense.
Yesterday i did enabled the other VLAN's and then i did get the disconnect status.
The only difference was that i had the VLAN's in the OC200 as interface.
I also forgot to disconnect the TL-ER605(router). So one of these are causing the disconnect.
Maybe this wil help you.
- Copy Link
- Report Inappropriate Content
If you set up VLAN and wrongly on the ports, this will disconnect the switch. If the switch shows up as "disconnected", you need to plug and cable or reboot them to recover. If the heartbeat is missed, this can auto recover if the connection is normal. Because the device keeps sending packets. If the controller detects no packets, a few tries later, that controller knows that specific device is totally offline. So, turn to disconnected. Unless the device power cycle or have connections again. https://www.tp-link.com/en/support/faq/2980/
Heartbeat missed:
A transition status between Connected and Disconnected. Once connected to the controller, the device will send inform packets to the controller in a regular interval to maintain the connection. If the controller does not receive its inform packets in 30 seconds, the device will turn into the Heartbeat Missed status. For a heartbeat-missed device, if the controller receives an inform packet from the device in 5 minutes, its status will become Connected again; otherwise, its status will become Disconnected
VLAN 1 is not really a "VLAN". It's just the default network before you start a virtual network. When you start VLAN 2 3 4..., they're real virtual networks.
If you don't have an Omada router, you don't use VLAN interface because VLAN interface you see on the Cotroller is an integrated all-in-one setup that requires the router. If you got switch, you don't have to set up specific settings because it's so easy to configure with APs with SSID VLAN. When it comes to regular PC or another switch, must got some knowledge about VLAN tag/untag. Or definitely will get trouble into setting up and wreck the whole network but to reset to recover.
If your router is another brand, you just need to set up the VLAN to make sure the ports are clear for the pass.
- Copy Link
- Report Inappropriate Content
Thanks for all replays,
I have solved the issue by setting al VLANs again on the switch, as I did not remember how I set the first time (a few month before that).
The main point - management VLAN must be untagged, that just means that the device (switch or AP) will get it's ip address from DHCP server as it's connected without any VLAN.
- Copy Link
- Report Inappropriate Content
Information
Helpful: 0
Views: 3124
Replies: 7
Voters 0
No one has voted for it yet.