|
|
This chapter describes maintenance procedures that should be performed regularly to keep the telephony controller operations running smoothly. These maintenance procedures include:
The telephony controller software includes a disk monitoring program that periodically checks the amount of disk space used within the configurable set of disk partitions. This ensures that there is sufficient disk space available in each disk partition for the telephony controller to continue to operate at peak performance.
If any disk partition exceeds the configurable usage threshold, the telephony controller generates a DISK alarm, a warning of a disk partition overrun, and a warning of insufficient disk space. Because the DISK alarm is a major alarm, you need to clear it as soon as possible. See the "Clearing Platform Alarms" section for more information about the DISK alarm.
In addition to the DISK alarm, the program runs a script that deletes log files older than a pre-defined time span in the /opt/TransPath/var/log and /opt/TransPath/var/spool directories until the disk space usage is within the designated threshold.
The disk monitor is installed with the default value of 80% for the usage threshold. When disk usage exceeds 80% of the disk capacity, the DISK alarm is generated.
However, you can configure the disk monitor program to match your site-specific requirements. To configure the program, perform the following steps:
Step 1 Log in to the telephony controller as Root.
Step 2 Change to the /opt/TransPath/local directory.
Step 3 Use a text editor to open the diskmonitor.sh program.
Step 4 Locate the customizable variables section in the beginning of the file. This section will look similar to the following sample screen output:
######### Customizable variables ############ # Add the files systems to check here filesys="/opt/TransPath" # Change the threshold to alarm upon here MAX_FULL=80 # Change the number of days back to begin trimming here MAX_DAYS=7
Step 5 Change the MAX_FULL parameter to match your site-specific requirements.
The MAX_FULL parameter sets the threshold for generating a DISK alarm. The units of disk space in use is expressed as a percentage of the amount of available partitioned disk space. For example, in the file section shown above, the MAX_FULL parameter is set to 80. Therefore, the DISK alarm is generated when disk usage exceeds 80% of capacity.
In addition, periodically back up the full system to capture information about the setup and configuration of the telephony controller. In case of a catastrophic failure, it is much easier to restore system information from backup data than to recreate it. Furthermore, in case of such a failure, critical configuration information may be lost if it has not been backed up. You can use either a UNIX tar or ufsdump command to perform these types of backups.
Table 5-1 provides a brief overview of backup methods and their benefits and limitations.
| Backup method | Recommended for | Pros | Cons |
|---|---|---|---|
Telephony controller backup utility | Regular backups of telephony controller data to tape. | Uses one easy command to back up data. |
|
UNIX tar command | Full-system backups to various media. |
| Slow if restoring individual files, rather than complete directory. |
UNIX ufsdump command | Full-system backups to various media. The preferred utility for full system backups. |
| Backs up only single file systems, so you must enter multiple ufsdump commands to back up data. |
Complete data backups should be done once a week. Perform these backups during the least busy time periods (for example, weekends or evenings). Backups of system configuration information should be made on an as-needed basis. For example, you may want to create a backup after any major system configuration changes.
If you have two telephony controllers in a failover configuration, Cisco does not recommend that you back up one system to the other. This method wastes disk space that is normally used for spool and log files on the secondary system.
To back up your system using the backup utility, perform the following steps:
Step 1 Log in to the telephony controller as a user in the transpath group and enter the following command:
backup.sh
The system returns a message similar to the following:
TransPath 2000 backup utility
-----------------------------
Enter:
<F> for Full (everything you have)
<P> for Partial (changeable part of the system)
<I> for Partial Incremental (everything changed since last P)
<Q> to quit
Select backup mode:
Step 2 Enter F, P, or I depending on your needs. The menu choices are:
The backup utility is automatically configured for a standard telephony controller site. However, if you receive error messages, you may need to modify the $TAPE_DEVICE variable in the backup.sh file.
To modify the $TAPE_DEVICE variable, perform the following steps:
Step 1 Log in to the telephony controller and change to the /opt/TransPath/local directory.
Step 2 Use a text editor to change the $TAPE_DEVICE variable in the file to the name of your tape device.
Use the UNIX tar command to back up your entire system. This method is easy to use for full backups, because it requires only a single command. Use this method if you do not expect to have to restore files often, or if you do not need to restore files quickly. If you use this method and later need to restore individual files (rather than the entire system), it will take longer to locate and restore the files.
To perform a backup using the tar command, log in to the telephony controller and enter the following:
tar cf device name
The device name parameter specifies the device to which you are backing up data: tape drive, CD-ROM, file, or diskette.
The UNIX ufsdump command is the best backup method for backing up your entire system. The ufsdump command allows you to back up file systems to tape, CD-ROM, a diskette, or a file.
To perform a backup using the ufsdump command, enter the following command for each file system:
ufsdump leveluf devicename filesystem
The level parameter is a number (0-9) that you can use to perform incremental backups. Cisco recommends that you perform full backups and set the level to 0. If you perform incremental backups, you designate a level each time a backup is done. This level is written to the dump dates file in your system. The next time you perform a backup, the system backs up everything that has changed since the last dump level. For example, if you entered a level of 5, the system would back up everything that had changed since level 4.
The devicename parameter specifies the device to which you are backing up data: tape drive, CD-ROM, file, or diskette.
The filesystem parameter specifies the files that are to be backed up. To see a list of all file systems, enter the following command:
df -lk
The system returns a message similar to the following:
Filesystem kbytes used avail capacity Mounted on /dev/dsk/c0t0d0s0 245980 12004 209386 6% / /dev/dsk/c0t0d0s5 984435 109921 776074 13% /usr /proc 0 0 0 0% /proc fd 0 0 0 0% /dev/fd /dev/dsk/c0t0d0s3 245980 3438 217952 2% /var /dev/dsk/c0t0d0s6 425743 9 383164 1% /export/home /dev/dsk/c0t0d0s4 1538867 9539 1375448 1% /opt swap 606552 152 606400 1% /tmp
Cisco recommends that you back up any file system with a device name. In the previous example, you would back up the following file systems:
/dev/dsk/c0t0d0s0 245980 12004 209386 6% / /dev/dsk/c0t0d0s5 984435 109921 776074 13% /usr /dev/dsk/c0t0d0s3 245980 3438 217952 2% /var /dev/dsk/c0t0d0s6 425743 9 383164 1% /export/home /dev/dsk/c0t0d0s4 1538867 9539 1375448 1% /opt
 | Caution Back up each file system to separate media (for example, use a separate CD-ROM for each file system). This prevents the ufsdump command from overwriting previous backups each time it creates a backup. For tape backup devices, use separate tapes or a "no-rewind" command to store multiple backups sequentially on tape. To write to the tape in no-rewind mode, enter the following command: |
dev/rmt/number n
The number parameter specifies the number of the tape drive connected to your hardware.
| Caution If you experience a failure that requires backup files to be restored, it is important that you first contact the Cisco Systems Technical Assistance Center (TAC) for specific instructions. Performing the restore process can result in key files being overwritten and possibly affect call processing. |
U.S. customers can reach the TAC at 1-800-553-2447. International customers should dial +1-408-526-7209. The TAC is also online at www.cisco.com. The e-mail address for the TAC is tac@cisco.com. See the "Contacting the TAC" section for more details about the TAC.
Posted: Fri Apr 30 15:25:21 PDT 1999
Copyright 1989-1999©Cisco Systems Inc.