Overview
Since Version 6 is always connected to a server, it is possible to collect information
through this connection in
case of problems.
The Online Information Collection functionality enables you to transfer the collected information directly to an
EV6 server. You will have to enable the functionality to send collected
information back to its server.
Currently, a Version 6 session generates a set of files: different statistics
files, abend traces and potentially dump files. They are stored in
different places and accumulate from one session to another.
To enable
the automatic collection of all information, the existing files need to
be packed together on a session basis in a zip file.
A Few Definitions
- Information
- Data allowing you to understand, visualize, analyze, calculate,
and estimate something. In this context, the information mentioned
is all
the data needed to understand a runtime problem, the scenario
leading to this problem, and the environment in which this problem occurs.
The information will enable you to analyze performance issues and not only crashes,
and also enable
all statistics and performance logs that a session can produce
- Abend Trace
- A text file containing information on the location of a crash
but also information on the history of a session and the
environment.
- Dump
- A binary file containing a snapshot of the process memory and
usable in a debugger.
- Statistics
- Files collecting specific user information describing the behavior of the application. Currently, there are PCS, command,
workbench, session, access and error statistics.
Activate
Check the Activate option to activate online information
collection.
By default, the online information collection (OIC) feature packages
the current logs that are generated in the CATTemp directory for each
session error.log, SessionInfo file and the abend trace file, in case of
an application crash.
The OIC does not activate the statistics directly. However, when it
is active the statistic files are packaged with the other information.
The activation of other specific statistics will add one file per
statistic to the previous set of data.
In this first step, form an end user point of view, no new data will
be generated. The only difference will reside in their recollection
procedure.
In the second step (correlated to a modification of the error
reporting process), an MS mini-dump could be added to the previous set
of files.
- HTTP Server URL
-
Specify the HTTP Server URL of the server to perform
online information collection.
It must be an existing http server that you have already deployed. By
default, the OIC feature uses a PUT request to send the zip archive. The
server has to allow the PUT request on the defined server.
- Use POST request
-
Check the Use POST request option if you do not want to
authorize PUT requests on the server.
If you do not want to allow the PUT requests, you can deploy an
applet or a cgi-bin script responding to POST requests that will store
the archives.