The fjsnap or pclsnap command is a tool which gathers system information necessary for analyzing the trouble at a time. When the trouble occurs by the PRIMECLUSTER system, the cause can be pursued by collecting necessary error information by the fjsnap or pclsnap command.
You can execute this command as follows:
Log in with the system administrator authority.
Execute the "fjsnap" or "pclsnap" command.
For fjsnap
/usr/sbin/fjsnap -a output
For pclsnap
/opt/FJSVpclsnap/bin/pclsnap -a output
The file name which becomes an output destination of system information collected by using the fjsnap or pclsnap command for output is specified.
The following messages may be output to a switchlog and /var/log/messages when the fjsnap or pclsnap command is executed while one or more cluster nodes are stopped. However, no action is required for these messages.
(BM, 8) Failed sending message <message> to object <object> on host <host>.
(WRP, 11) Message send failed, queue id <queueid>, process <process>, <name>, to host <node>.
See
For details on the "fjsnap" command, see the "README" file included in the "FJSVsnap" package.
For details on the "pclsnap" command, see the "README" file included in the "FJSVpclsnap" package.
Information
Execution timings for the fjsnap or pclsnap command
For problems that occur during operation, for example, if an error message is output, execute the "fjsnap" or "pclsnap" command immediately after the problem occurs.
If the "fjsnap" or "pclsnap" command cannot be executed because the system hangs, collect a crash dump. Then start the system in single user mode, and execute the "fjsnap" or "pclsnap" command.
For information on how to collect a crash dump, see "C.1.3 Crash Dump."
After an error occurs, if a node restarts automatically (the node could not be started in single-user mode) or if the node is mistakenly started in multi-user mode, execute the "fjsnap" or "pclsnap" command.
If investigation information cannot be collected because the "fjsnap" or "pclsnap" command results in an error, or the "fjsnap" or "pclsnap" command does not return, then collect a crash dump.
Free space in the directory required to run the fjsnap command
Check the README file for the fjsnap command for the size of the output directory and working directory (work area) required to run the fjsnap command. If there is not enough free space, reserve free space or change to a directory with a large amount of free space, and then execute the command.
In addition to the output directory and working directory listed in the README file for the fjsnap command, PRIMECLUSTER uses the following directories. Check the free space in the following directory. If there is not enough free space, reserve some free space before executing the command.
Under /opt/SMAW/SMAWRrms
Disk space:
Temporarily use the following directories and files. Use the du command to calculate and estimate the size of the following directories and files.
/var/opt/SMAWRrms/log
/var/log/messages*
Under /var/opt/FJSVcluster/data/TRC
Disk space:
About 2MB
Under /var/opt/FJSVsfcfs/log
Disk space:
The GFS Shared File System to use
Per 1 piece: 11.3 MB (Fixed)
Maximum (10 pieces): 113 MB (Fixed)
Under /tmp
Disk space:
The GFS Shared File System to use
Per 1 piece: 34.6 KB (Maximum)
Maximum (10 pieces): 346 KB (Maximum)
