Troubleshooting guide

3-24
Cisco PGW 2200 Softswitch Release 9.8 Operations, Maintenance, and Troubleshooting Guide
OL-0800-14
Chapter 3 Cisco PGW 2200 Softswitch Platform Operations
Periodic Maintenance Procedures
Automatic Disk Space Monitoring
The Cisco PGW 2200 Softswitch software includes a script that is called disk monitor (diskmonitor.sh).
Disk monitor periodically checks the amount of disk space that is used within the configurable set of
disk partitions. Disk monitor ensures that there is sufficient disk space available in each disk partition
for the system to continue to operate at peak performance. Disk monitor deletes (trims) the older log files
in the /opt/CiscoMGC/var/log and /opt/CiscoMGC/var/spool directories until the disk space usage is
within the specified threshold (set using the XECfgParm.dat parameter, diskmonitor.Threshold).
The disk monitor can also track the number of configurations that are stored in the configuration library
(which is found in the /opt/CiscoMGC/etc/CONFIB_LIB directory). Disk monitor trims the older
configurations when the number of configurations exceeds the maximum value you set in the associated
XECfgpParm.dat disk monitor parameter. The process manager runs the disk monitor script once every
minute.
For prior releases, users were encouraged to ensure that no more than 64 configurations were stored in
the configuration library. This recommendation was made to ensure the proper functioning of
switchovers and the prov-sync MML command. If the system stores more than 64 configurations,
switchovers and the prov-sync command time out and fail. Such timeouts and failures occur because the
standby Cisco PGW 2200 Softswitch attempts to copy over all the configurations that are stored on the
active Cisco PGW 2200 Softswitch.
Currently, the Cisco PGW 2200 Softswitch software automatically manages the process that administers
the configuration library. You set the disk monitor parameter to establish the maximum number of
configurations that are allowed in the configuration library. The system trims the older configurations as
necessary.
The following parameters in the XECfgParms.dat file control the disk monitor:
• diskmonitor.Limit—Specifies the number of days to preserve data before the system trims data. The
default value is 7.
• diskmonitor.OptFileSys—Allows monitoring of optional user-configurable file systems. This utility
monitors the /opt file system for threshold crossing. Using this parameter, you can monitor additional file
systems (disk slices) by setting the parameter to the preferred directory, such as /tmp, /usr or /var. The
messages that are associated with this parameter are sent to the platform.log file. To retrieve these
messages, you must scan the platform.log file for messages using the following format:
Filesystem file_system_name has exceeded num percent full.
The following sample log entry shows how a log entry might appear:
Filesystem /var has exceeded 80 percent full
Note Disk monitor does not trim files in these file systems.
• diskmonitor.Threshold—Specifies the percentage of disk usage that, when exceeded, generates a
DISK alarm, and initiates disk trimming. The default value is 80.
• diskmonitor.CdrRmFinished—Specifies how many days to keep completed (polled) call detail
record (CDR) files. The default value is 0. The default specifies that the Cisco BAMS polls the Cisco
PGW 2200 Softswitch. The CDR.bin files remain in a user-configurable directory until the Cisco BAMS
trims them (using the format CDR_timestamp.finished). Alternatively, the disk monitor trims the file
from the user-configurable directory.