Routing

Our school district has A/V devices in every classroom. I sometimes get a report that a device can not be found with a ping. I have found that if I do a ping sweep of the whole school IP range for these devices, I do not ret a response from a device.  If I ping from the CLI of the connected switch it also fails. If I ping from the core switch, which is the router for the school and is also where the interface irp is defined, it does respond.  If I clear the ARP table on the core switch and ping from outside the school, I get the same result and still can not ping it from the connected switch or any other switch but the core switch.  

I can't think of a reason why this would happen. 

This is likely related to how the vlans are setup with the gateways on the core switch,  and the vlan where the mgmt interface is on the edge switch where the device is connected.

When you ping from the edge switch this will be sourced from the local mgmt ip with a destination of the ip address of the device.  This will need to route via layer 3.  So it appears there is no layer 3 path between those two vlans in at least one direction.

A diagram of the vlan layout would help to see where the potential routing problem exists.

Or running a traceroute from the edge and also from the device to the edge switch ip would help to see where the path has an issue.

Our school district has A/V devices in every classroom. I sometimes get a report that a device can not be found with a ping. I have found that if I do a ping sweep of the whole school IP range for these devices, I do not ret a response from a device.  If I ping from the CLI of the connected switch it also fails. If I ping from the core switch, which is the router for the school and is also where the interface irp is defined, it does respond.  If I clear the ARP table on the core switch and ping from outside the school, I get the same result and still can not ping it from the connected switch or any other switch but the core switch.  

I can't think of a reason why this would happen. 

This is likely related to how the vlans are setup with the gateways on the core switch,  and the vlan where the mgmt interface is on the edge switch where the device is connected.

When you ping from the edge switch this will be sourced from the local mgmt ip with a destination of the ip address of the device.  This will need to route via layer 3.  So it appears there is no layer 3 path between those two vlans in at least one direction.

A diagram of the vlan layout would help to see where the potential routing problem exists.

Or running a traceroute from the edge and also from the device to the edge switch ip would help to see where the path has an issue.

------------------------------
Steve Puluka BSEET - Juniper Ambassador
IP Architect - DQE Communications Pittsburgh, PA (Metro Ethernet & ISP - Retired)
http://puluka.com/home

Original Message:
Sent: 09-13-2024 12:05
From: JOHN WILLIAMSON
Subject: A device can be pinged from core switch ( router) but not the switch it's plugged into. Why?

Our school district has A/V devices in every classroom. I sometimes get a report that a device can not be found with a ping. I have found that if I do a ping sweep of the whole school IP range for these devices, I do not ret a response from a device.  If I ping from the CLI of the connected switch it also fails. If I ping from the core switch, which is the router for the school and is also where the interface irp is defined, it does respond.  If I clear the ARP table on the core switch and ping from outside the school, I get the same result and still can not ping it from the connected switch or any other switch but the core switch.  

I can't think of a reason why this would happen. 

------------------------------
JOHN WILLIAMSON
------------------------------

This is likely related to how the vlans are setup with the gateways on the core switch,  and the vlan where the mgmt interface is on the edge switch where the device is connected.

When you ping from the edge switch this will be sourced from the local mgmt ip with a destination of the ip address of the device.  This will need to route via layer 3.  So it appears there is no layer 3 path between those two vlans in at least one direction.

A diagram of the vlan layout would help to see where the potential routing problem exists.

Or running a traceroute from the edge and also from the device to the edge switch ip would help to see where the path has an issue.

Our school district has A/V devices in every classroom. I sometimes get a report that a device can not be found with a ping. I have found that if I do a ping sweep of the whole school IP range for these devices, I do not ret a response from a device.  If I ping from the CLI of the connected switch it also fails. If I ping from the core switch, which is the router for the school and is also where the interface irp is defined, it does respond.  If I clear the ARP table on the core switch and ping from outside the school, I get the same result and still can not ping it from the connected switch or any other switch but the core switch.  

I can't think of a reason why this would happen. 

I have found the mac address is showing on the edge switches and will show in the arp table on the core switch. 

This is likely related to how the vlans are setup with the gateways on the core switch,  and the vlan where the mgmt interface is on the edge switch where the device is connected.

When you ping from the edge switch this will be sourced from the local mgmt ip with a destination of the ip address of the device.  This will need to route via layer 3.  So it appears there is no layer 3 path between those two vlans in at least one direction.

A diagram of the vlan layout would help to see where the potential routing problem exists.

Or running a traceroute from the edge and also from the device to the edge switch ip would help to see where the path has an issue.

Our school district has A/V devices in every classroom. I sometimes get a report that a device can not be found with a ping. I have found that if I do a ping sweep of the whole school IP range for these devices, I do not ret a response from a device.  If I ping from the CLI of the connected switch it also fails. If I ping from the core switch, which is the router for the school and is also where the interface irp is defined, it does respond.  If I clear the ARP table on the core switch and ping from outside the school, I get the same result and still can not ping it from the connected switch or any other switch but the core switch.  

I can't think of a reason why this would happen. 

I could still use some help understanding this issue.  In a nut shell, why can I ping a device at a school on the core switch but not on the switch the device is plugged into or anywhere else in our school district? What could be different about this service, if almost all of the rest of them work as expected?

I have found the mac address is showing on the edge switches and will show in the arp table on the core switch. 

This is likely related to how the vlans are setup with the gateways on the core switch,  and the vlan where the mgmt interface is on the edge switch where the device is connected.

When you ping from the edge switch this will be sourced from the local mgmt ip with a destination of the ip address of the device.  This will need to route via layer 3.  So it appears there is no layer 3 path between those two vlans in at least one direction.

A diagram of the vlan layout would help to see where the potential routing problem exists.

Or running a traceroute from the edge and also from the device to the edge switch ip would help to see where the path has an issue.

Our school district has A/V devices in every classroom. I sometimes get a report that a device can not be found with a ping. I have found that if I do a ping sweep of the whole school IP range for these devices, I do not ret a response from a device.  If I ping from the CLI of the connected switch it also fails. If I ping from the core switch, which is the router for the school and is also where the interface irp is defined, it does respond.  If I clear the ARP table on the core switch and ping from outside the school, I get the same result and still can not ping it from the connected switch or any other switch but the core switch.  

I can't think of a reason why this would happen. 

Since the trace route dies at the core there is a reachability issue between the mgmt vlan and vlan 93 in this case.

Are all gateways for the vlans on the core switch?

Is the mgmt vlan gateway on the core switch the default route destination configured on the edge switches?

Do all devices on vlan 93 use the gateway on the core switch as their gateway or default route?

I could still use some help understanding this issue.  In a nut shell, why can I ping a device at a school on the core switch but not on the switch the device is plugged into or anywhere else in our school district? What could be different about this service, if almost all of the rest of them work as expected?

I have found the mac address is showing on the edge switches and will show in the arp table on the core switch. 

This is likely related to how the vlans are setup with the gateways on the core switch,  and the vlan where the mgmt interface is on the edge switch where the device is connected.

When you ping from the edge switch this will be sourced from the local mgmt ip with a destination of the ip address of the device.  This will need to route via layer 3.  So it appears there is no layer 3 path between those two vlans in at least one direction.

A diagram of the vlan layout would help to see where the potential routing problem exists.

Or running a traceroute from the edge and also from the device to the edge switch ip would help to see where the path has an issue.

Our school district has A/V devices in every classroom. I sometimes get a report that a device can not be found with a ping. I have found that if I do a ping sweep of the whole school IP range for these devices, I do not ret a response from a device.  If I ping from the CLI of the connected switch it also fails. If I ping from the core switch, which is the router for the school and is also where the interface irp is defined, it does respond.  If I clear the ARP table on the core switch and ping from outside the school, I get the same result and still can not ping it from the connected switch or any other switch but the core switch.  

I can't think of a reason why this would happen. 

Do all devices on vlan 93 use the gateway on the core switch as their gateway or default route?

Yes.  Or at least, they were all supposed to.  We had a vendor install an A/V system in every classroom in the school district and I gave them the network info to use. 

Do all devices on vlan 93 use the gateway on the core switch as their gateway or default route?

Yes

Are all gateways for the vlans on the core switch?

Yes

The annoying part is that all of these devices did work at some point.  Otherwise, they could not have been programmed.  We have seen devices show up in the system and then drop off.  Of course, the installing company blames the network. 

Since the trace route dies at the core there is a reachability issue between the mgmt vlan and vlan 93 in this case.

Are all gateways for the vlans on the core switch?

Is the mgmt vlan gateway on the core switch the default route destination configured on the edge switches?

Do all devices on vlan 93 use the gateway on the core switch as their gateway or default route?

I could still use some help understanding this issue.  In a nut shell, why can I ping a device at a school on the core switch but not on the switch the device is plugged into or anywhere else in our school district? What could be different about this service, if almost all of the rest of them work as expected?

I have found the mac address is showing on the edge switches and will show in the arp table on the core switch. 

This is likely related to how the vlans are setup with the gateways on the core switch,  and the vlan where the mgmt interface is on the edge switch where the device is connected.

When you ping from the edge switch this will be sourced from the local mgmt ip with a destination of the ip address of the device.  This will need to route via layer 3.  So it appears there is no layer 3 path between those two vlans in at least one direction.

A diagram of the vlan layout would help to see where the potential routing problem exists.

Or running a traceroute from the edge and also from the device to the edge switch ip would help to see where the path has an issue.

Our school district has A/V devices in every classroom. I sometimes get a report that a device can not be found with a ping. I have found that if I do a ping sweep of the whole school IP range for these devices, I do not ret a response from a device.  If I ping from the CLI of the connected switch it also fails. If I ping from the core switch, which is the router for the school and is also where the interface irp is defined, it does respond.  If I clear the ARP table on the core switch and ping from outside the school, I get the same result and still can not ping it from the connected switch or any other switch but the core switch.  

I can't think of a reason why this would happen. 

Are the devices DHCP or static addressed?

If they are DHCP perhaps they are failing in lease renewals and that is why they drop off apparently randomly.

Or perhaps the devices in low power mode or similar are bouncing the ports and the dynamic port configuration is also bouncing or changing.  Check logs for messages in any known outage windows.

Do all devices on vlan 93 use the gateway on the core switch as their gateway or default route?

Yes.  Or at least, they were all supposed to.  We had a vendor install an A/V system in every classroom in the school district and I gave them the network info to use. 

Do all devices on vlan 93 use the gateway on the core switch as their gateway or default route?

Yes

Are all gateways for the vlans on the core switch?

Yes

The annoying part is that all of these devices did work at some point.  Otherwise, they could not have been programmed.  We have seen devices show up in the system and then drop off.  Of course, the installing company blames the network. 

Since the trace route dies at the core there is a reachability issue between the mgmt vlan and vlan 93 in this case.

Are all gateways for the vlans on the core switch?

Is the mgmt vlan gateway on the core switch the default route destination configured on the edge switches?

Do all devices on vlan 93 use the gateway on the core switch as their gateway or default route?

I could still use some help understanding this issue.  In a nut shell, why can I ping a device at a school on the core switch but not on the switch the device is plugged into or anywhere else in our school district? What could be different about this service, if almost all of the rest of them work as expected?

I have found the mac address is showing on the edge switches and will show in the arp table on the core switch. 

This is likely related to how the vlans are setup with the gateways on the core switch,  and the vlan where the mgmt interface is on the edge switch where the device is connected.

When you ping from the edge switch this will be sourced from the local mgmt ip with a destination of the ip address of the device.  This will need to route via layer 3.  So it appears there is no layer 3 path between those two vlans in at least one direction.

A diagram of the vlan layout would help to see where the potential routing problem exists.

Or running a traceroute from the edge and also from the device to the edge switch ip would help to see where the path has an issue.

Our school district has A/V devices in every classroom. I sometimes get a report that a device can not be found with a ping. I have found that if I do a ping sweep of the whole school IP range for these devices, I do not ret a response from a device.  If I ping from the CLI of the connected switch it also fails. If I ping from the core switch, which is the router for the school and is also where the interface irp is defined, it does respond.  If I clear the ARP table on the core switch and ping from outside the school, I get the same result and still can not ping it from the connected switch or any other switch but the core switch.  

I can't think of a reason why this would happen.