What network related information should be collected when experiencing a problem with IBM FileNet P8 Content Management software?
Collecting essential data early, even before opening a PMR, helps IBM Software Support quickly determine the cause of the problem.
In addition to the information in the MustGather: Read first for FileNet Content Engine, gather the following network related information:
- Collect IP addresses of all servers in the IBM FileNet P8 Content Management environment
- A network topology diagram is required.
- For issues that consist of an error log entry, a network trace should be captured from the node where the error log resides
- If the Content Management environment consist of two servers or more, it is best to identify the source and destination servers of the communications and capture all traffic in and out of the identified servers.
- If the Content Management environment is configured with a Load Balancer or another type of High Availability device, include the IP addresses (Virtual IP addresses) of the device as well. In some cases capturing network traffic from the load balancer perspective is required. This should be discussed with an IBM Network specialist for clarification.
- If the performance/integrity issue can be reproduced at will, start the tracing tool (i.e. SnifferPro, Wireshark, Snoop, iptrace, nettl, etc...) and perform the actions to reproduce the issue. See the attached Packet Sniffer document for more detailed information on using the network capturing tools
- If the performance/integrity issue cannot be reproduced at will, identify a time frame when the issue typically occurs. Start the tracing tool (as noted above) during the identified time frame, and capture traffic until the issue occurs.
- As soon as you have reproduced the issue, stop the trace and save the data.
- Setting up filters or triggers to collect data is an option, but should be discussed with a IBM FileNet network analyst for proper configuration.
- Prepare the trace and any other collected data (ex: network topology and problem description, vwtrace logs, log4j, etc) to send to IBM by compressing the data into 100MB files using Unix "tar" or Windows Winzip.