This article describes the aspects which have to be taken into account for trace recording in order for the traces to be expressive and useful for error analysis.
- A trace always has to be recorded in the error state.
- The exact time has to be noted, when the error occured, so that the trace analysis can be focused on this time.
Specific requirements for certain topics:
- When connection problems between an ePaper access point and the ePaper server occur, the IP addresses of the access point and the server are necessary for analysis.
- If the VPN connection is between two LANCOM routers, the traces have to be recorded on both routers at the same time.
- The traces have to be recorded locally and not via the VPN connection as otherwise important information would be lost after a disconnect of the VPN connection!
- If multiple VPN connections are established on a router, the VPN-IKE trace has to be filtered to the public IP address of the other VPN member.
- The involved telephone numbers have to noted, so that the trace analysis can be focused on these.
- When the following problems occur an additional Wireshark trace has to be recorded of the Internet connection as well as from the local network, where the SIP user is located. The procedure is described in this Knowledge Base article.
- One-sided voice transmission
- Missing voice transmission
The Wireshark traces have to be recorded with the tool LCOScap as the latency with this tool is lower, than Wireshark traces recorded via WEBconfig!
- If the behaviour cannot be narrowed down to a single access point, the traces have to be recorded on all of the potentially affected access points.
- The MAC address of the affected WiFi device has to be entered as the Trace MAC.
- The MAC address of the affected WiFi device is necessary for the trace analysis.
WLAN Controller Scenario:
- The traces have to be recorded on the WLAN Controller and an affected access point at the same time in the error state.
- The CAPWAP-CTRL trace on the WLAN Controller has to be filtered to the IP address of an affected access point, so that the trace output is limited to this specific access point.