Weird trunk link issue
I don't know that this is a switch issue. Most likely it actually isn't.
My topology is:
Home PC ->TailScale VPN->Office Omada Gateway (ER707-M2)->Omada Switch (SL2428P)->Non Omada Switch (Zyxel GS1900)
I have a server at home that runs a script every hour and sends 1 ping to a variety of devices inside my network.
I have 1 device (the non Omada switch listed above) that sometimes doesn't respond on it's mngmt IP address. Here is where it gets weird though. Devices connected to that switch repsond just fine.
I have replaced all cabling except what is going through the ceiling (that's probably next)
I have replaced the switch (originally it was a Cisco SG350).
I have looked for a duplicate IP with no success.
Sometimes it will go a full day and never miss a ping cycle. sometimes it will miss 50%. To be clear, once I get the email that it is not responding, most of the time I can ping it manually and it doesn't respond. So it isn't just a momentary blip. But, I can ALWAYS ping devices hanging off the switch, so it's not a link failure between the Omada switch and the Zyxel. It seems to only be the VLAN. It is also ONLY this device that does this. The rest of the devices (some PCs, other switches, and various other devices) are responding as they should.
Any idea where to look? Like I said, I will probably replace, or at a minimum re-terminate the connections going through the ceiling. It just seems weird that the management VLAN doesn't respond when the untagged VLAN responds normally.
Also, one note. I am not sure if this only happens over the VPN, but my guess would be that the VPN isn't the issue since other devices respond just fine.
- Copy Link
- Subscribe
- Bookmark
- Report Inappropriate Content
Hi @muzicman0
Thanks for posting in our business forum.
Reading the first context, it gives me a feeling that the device has gone to sleep mode?
It is not a problem with the software/hardware and good to know that you have checked them. I am inclined to the view that it has gone to sleep mode or some kind of hibernation.
It does not matter what kind of switch/router you connect it to. It is not a problem with the networking devices and your design.
- Copy Link
- Report Inappropriate Content
@muzicman0 The switch that is being pinged doesn't have a sleep mode. Also, it will come back after some time. sometimes it is seconds, sometimes it is 30 minutes, but it always comes back with no interaction from anyone.
Also, just to rule out a duplicated IP address, I did change the address. it still had issues, so that is also one more thing it isn't.
- Copy Link
- Report Inappropriate Content
Hi @muzicman0
Thanks for posting in our business forum.
muzicman0 wrote
@muzicman0 The switch that is being pinged doesn't have a sleep mode. Also, it will come back after some time. sometimes it is seconds, sometimes it is 30 minutes, but it always comes back with no interaction from anyone.
Also, just to rule out a duplicated IP address, I did change the address. it still had issues, so that is also one more thing it isn't.
The issued device is a switch? The mentioned model, ours?
- Copy Link
- Report Inappropriate Content
@Clive_A as mentioned in the original post, it is a Zyxel GS1900, but originally it was a Cisco SG350. I replaced the Cisco with the Zyxel because of this issue, but the issue persists, so it's not the switch. Also as mentioned, PC's (edge devices) connected to the Zyxel respond just fine even when the switch doesn't respond. The IP address not responding is on the default VLAN (which is VLAN 254). The edge devices that ARE responding are on VLAN 108. But again, other devices on VLAN 254 are responding just fine and don't have issues.
- Copy Link
- Report Inappropriate Content
Hi @muzicman0
Thanks for posting in our business forum.
muzicman0 wrote
@Clive_A as mentioned in the original post, it is a Zyxel GS1900, but originally it was a Cisco SG350. I replaced the Cisco with the Zyxel because of this issue, but the issue persists, so it's not the switch. Also as mentioned, PC's (edge devices) connected to the Zyxel respond just fine even when the switch doesn't respond. The IP address not responding is on the default VLAN (which is VLAN 254). The edge devices that ARE responding are on VLAN 108. But again, other devices on VLAN 254 are responding just fine and don't have issues.
Remove the SL switch, and place it with Cisco in between, will it work?
The issue persists? If it does, it does not look like a problem with SL model.
- Copy Link
- Report Inappropriate Content
@Clive_A re-working my topology isn't an option right now, this is in production. However, I did replace the keystone jack in the office, and re-punch down the core switch side to the patch panel, and it has not had an error for a little over 24 hours, so I believe this is solved.
- Copy Link
- Report Inappropriate Content
@Clive_A It happened again. This was in the event log:
| hw-sung-jin-sw was disconnected from network "management_vlan(Default)" on hw-24poe-sw(connected time:45h3m connected, traffic: 0Bytes) and connected to network "sung_jin_vlan" on hw-24poe-sw. |
Sep 12, 2024 07:41:01 |
| hw-sung-jin-sw was disconnected from network "sung_jin_vlan" on hw-24poe-sw(connected time:45h6m connected, traffic: 0Bytes) and connected to network "management_vlan(Default)" on hw-24poe-sw. |
Sep 12, 2024 07:44:09 |
For some reason, the switch appears to be going from one VLAN to another, then switched back to the correct VLAN 3 minutes later. I don't understand how that could happen?
- Copy Link
- Report Inappropriate Content
Hi @muzicman0
Thanks for posting in our business forum.
muzicman0 wrote
@Clive_A It happened again. This was in the event log:
hw-sung-jin-sw was disconnected from network "management_vlan(Default)" on hw-24poe-sw(connected time:45h3m connected, traffic: 0Bytes) and connected to network "sung_jin_vlan" on hw-24poe-sw.
Sep 12, 2024 07:41:01
hw-sung-jin-sw was disconnected from network "sung_jin_vlan" on hw-24poe-sw(connected time:45h6m connected, traffic: 0Bytes) and connected to network "management_vlan(Default)" on hw-24poe-sw.
Sep 12, 2024 07:44:09
For some reason, the switch appears to be going from one VLAN to another, then switched back to the correct VLAN 3 minutes later. I don't understand how that could happen?
That might not be accurate.
What's the result of disabling the VLAN to see if this persists? I don't think VLAN would be a problem in this situation.
- Copy Link
- Report Inappropriate Content
Information
Helpful: 0
Views: 361
Replies: 8
Voters 0
No one has voted for it yet.