Collecting information for troubleshooting: Difference between revisions

From OPC Labs Knowledge Base
(→‎Preparations: fix link)
No edit summary
Line 3: Line 3:


== Preparations ==
== Preparations ==
Find the simplest case that shows first symptoms of a problem, and use that in a test. The logs are tedious to analyze even if they are short; any unnecessary communication adds tremendously to the complexity of the investigation.


# For OPC Classic, install OPC Analyzer. For OPC UA, install Wireshark.
# For OPC Classic, install OPC Analyzer. For OPC UA, install Wireshark.
Line 9: Line 10:
# Instrument your code so that it collects and logs relevant application-specific occurrences, together with their timestamps and other details.
# Instrument your code so that it collects and logs relevant application-specific occurrences, together with their timestamps and other details.
# Enable logging on the side of the OPC server, if possible.
# Enable logging on the side of the OPC server, if possible.
== Learning material ==
* [[How to use OPC Analyzer]]
* Analyzing OPC UA Communications with Wireshark: http://opcconnect.opcfoundation.org/2017/02/analyzing-opc-ua-communications-with-wireshark/
* OPC UA & Wireshark: https://www.prosysopc.com/blog/opc-ua-wireshark/


== Assets to collect ==
== Assets to collect ==
Line 18: Line 24:
# A log of relevant occurrences in your own application.
# A log of relevant occurrences in your own application.
# OPC server log(s), if available.
# OPC server log(s), if available.
It is important that all logs contain timestamps (with milliseconds precision, or at least the highest precision achievable), so that we can correlate them properly.


In addition, please include any relevant observations and explanations as to what has been attempted and how the test ran.
In addition, please include any relevant observations and explanations as to what has been attempted and how the test ran.

Revision as of 09:21, 24 May 2017

This article describes the assets that you need to collect for a successful "deep" troubleshooting.

Preparations

Find the simplest case that shows first symptoms of a problem, and use that in a test. The logs are tedious to analyze even if they are short; any unnecessary communication adds tremendously to the complexity of the investigation.

  1. For OPC Classic, install OPC Analyzer. For OPC UA, install Wireshark.
  2. For OPC UA, enable extended tracing (see QuickOPC-UA: How to enable extended tracing). Configure it so that the information is logged to a file, and contains timestamps and other details.
  3. For OPC UA, instrument your code so that EasyUAClient.LogEntry notifications are logged to a file, together with their timestamps and other details.
  4. Instrument your code so that it collects and logs relevant application-specific occurrences, together with their timestamps and other details.
  5. Enable logging on the side of the OPC server, if possible.

Learning material

Assets to collect

We will need following files from you:

  1. Trace files from OPC Analyzer (for OPC Classic) or Wireshark (for OPC UA).
  2. Extended tracing output (for OPC UA only).
  3. EasyUAClient.LogEntry records (for OPC UA only).
  4. A log of relevant occurrences in your own application.
  5. OPC server log(s), if available.

It is important that all logs contain timestamps (with milliseconds precision, or at least the highest precision achievable), so that we can correlate them properly.

In addition, please include any relevant observations and explanations as to what has been attempted and how the test ran.