Path states definitions between the VMware SD-WAN Edges and/or VMware SD-WAN Gateways.
Environment
VMware SD-WAN by VeloCloud
Cause
An Administrator or Operator can confirm the status of a path between a Gateway or Edge on the Orchestrator's Remote Diagnostics page under "List Paths". This can be useful when troubleshooting connectivity issues.
Example:
Resolution
Note: This article does not aim to explain the different statuses displayed on the Monitoring tab of the Edge. This is specifically for the "List Paths" section of remote diagnostics and states seen within the logs. This is used when troubleshooting connectivity issues.
It's important to understand heartbeats in order to understand Path States. A heartbeat is a message sent to the edge's peer to verify the tunnel status. A Heartbeat (HB) interval is 100 ms when high priority traffic is flowing through the tunnel. When there is no high priority traffic the HB interval is changed to 500 ms.
There are 7 different types of SD-WAN Path States:
INITIAL: This is the first stage of a path. In this state, the edge sends a packet to the peer to initiate a VCMP tunnel. Once the peer receives the initial packet, it will reply with an acknowledgement. Once the Edge receives the acknowledge, it will reply with a "Done" message and transition to the STABLE state.
If a path is stuck in INITIAL, it could mean a connectivity issue between both devices. Check that both the edge and gateway are reachable and can talk to each other. Verify that UDP port 2426 is not being blocked by a firewall or service provider.
STABLE: This state means that the link to the peer has been established and bi-directional connectivity is possible. Link quality is within the SLA values. This is the optimal state all paths should be.
UNSTABLE: This means the path is currently established and bi-directional connectivity is possible, but the quality of the link is outside the SLA values. 10% Packet loss or 3 missed heartbeats (300 ms silence).
If a path is often found in this state, it could be a sign that the underlay connection to the gateway is having issues. The Service Provider should be engaged to isolate any issues with the physical connection.
QUIET: When a path that is STABLE does not receive any packets after 7 Heartbeats (700 ms silence), it is marked as quiet. This includes heartbeats to the peer.
If a path is often found as quiet, it could be a sign of packet loss. If this is the case, the UNSTABLE path state may also be seen from time to time.
STANDBY: This means the path is in a stable state, however, it was configured to be a back up link and will not pass user data traffic. Heartbeats and SD-WAN traffic will still egress through this path even when in Standby.
DEAD: The link has not received any packets after 70 Heartbeats (7 seconds silence). This means the path will be removed. If the path is still configured, it will then go back to INITIAL state to attempt to reestablish the link.
A link that constantly is marked as DEAD could mean connectivity issues between the edge and the peer. Check any intermediate devices between edge and peer. Engage the service provider to help isolate any underlay issues.