This document provides guidance on material to include with OCSS7 support requests.
Gathering Support Data
OCSS7 support requests should contain the following information:
-
An SGC Report for each SGC node in the SGC cluster
-
OCSS7 TCAP Stack Logging for each Rhino node in the Rhino cluster
Gathering Support Data from the SGC
The SGC installation includes a convenience script $SGC_HOME/bin/generate-report.sh that should be used to gather information for a support request. The information gathered by this script includes:
-
Information about the SGC version being used.
-
General system information, such as network interfaces, running processes, mounted filesystems.
-
The SGC’s general configuration files:
$SGC_HOME/config/* -
The SGC startup scripts:
$SGC_HOME/bin/* -
The SGC’s log files:
$SGC_HOME/logs/* -
The current state of the SGC as indicated by the various
display-*command line client commands, if the SGC is running. -
A thread dump from the SGC process, if the SGC is running.
-
The current saved MML configuration:
$SGC_HOME/var/sgc.dat
This is the preferred method of gathering information for a support request, and this script is considered safe to use on a live SGC.
Configuring the generate-report.sh script
The generate-report.sh script is configured with some default values that may require changing for non-standard installations.
These values may be changed by opening the script in your preferred text editor and following the instructions in that script.
|
Do not edit anything after the section marked "End of user-configurable section". |
| Property | What it specifies | Default |
|---|---|---|
|
The level of logging that the 0 = silent, 1 = basic information (default), 2 = debug, 3 = more debug |
1 |
|
Where the SGC is installed. |
The parent directory of the |
|
Where the SGC’s ss7.log and startup.log are saved. |
|
|
The JMX hostname to use to connect to the SGC to gather runtime status information. |
Auto-detected from |
|
The JMX port to use to connect to the SGC to gather runtime status information. |
Auto-detected from |
|
How many days of SGC logging to retrieve.
The script is conservative and will always ensure that there is at least 1GB or 10% (whichever is the larger) disk space remaining after gathering logs. |
|
|
The tar command to use to package up the generated report. By default |
|
Executing the generate-report.sh script
One node at a time, for each node in the cluster, execute the script (this may take a few minutes):
cd $SGC_HOME ./bin/generate-report.sh
Example output where LOGLEVEL is set to 1 (default) and the SGC is running looks like:
cath@badak:~/testing/PC2-1$ ./bin/generate-report.sh
Initializing...
- Using SGC_HOME /home/cath/testing/PC2-1 (if this is not correct, override in generate-report.sh)
Generating report...
- Getting general system information
- Getting info from /proc subsystem for SGC pid 28229
- Getting configuration files
- Getting SGC scripts
- Getting SGC logs
tar: Removing leading `/' from member names
- Getting runtime state from JMX
- Getting thread dump from SGC PC2-1 (pid=28229)
- Getting runtime SGC configuration
Cleaning up...
Report written to /home/cath/testing/PC2-1/ocss7-report-PC2-1-2017-12-18_103720.tar
*** Note that this report is not compressed. You may compress it with bzip2 or xz if you wish. ***
If the SGC is not running, the output may look like this:
cath@badak:~/testing/PC1-1$ ./bin/generate-report.sh Initializing... - Using SGC_HOME /home/cath/testing/PC1-1 (if this is not correct, override in generate-report.sh) [WARN]: The SGC is not running, limited reports will be generated. Generating report... - Getting general system information - Not getting /proc/ info for process (SGC is not running) - Getting configuration files - Getting SGC scripts - Getting SGC logs tar: Removing leading `/' from member names - Not getting runtime state from JMX (SGC is not running) - Not getting thread dump (SGC is not running) - Getting runtime SGC configuration Cleaning up... Report written to /home/cath/testing/PC1-1/ocss7-report-PC1-1-2017-12-18_090911.tar *** Note that this report is not compressed. You may compress it with bzip2 or xz if you wish. ***
|
As indicated by the script, it may be desirable to compress the resulting tar file.
We recommend using nice if performing this on the SGC host so as not to compromise node performance.
|
Please verify the contents of this file before uploading, especially if SGC components are in non-standard locations.
In particular, ensure that ss7.log- and startup.log- files are present.
Gathering Support Data from the OCSS7 TCAP Stack
The OCSS7 TCAP stack is an integral component of the CGIN RA. Please refer to Appendix C. Support Requests in the CGIN documentation for further details.
