cdm Advanced Configuration

This article describes the advanced configuration concepts and procedures for the CPU, Disk, Memory Performance Monitoring (cdm) probe.
This article describes the advanced configuration concepts and procedures for the CPU, Disk, Memory Performance Monitoring (cdm) probe.
Copy Probe Configuration Parameters
If you want similar cdm probe configuration across multiple robots, copy the configuration from one probe to another.
Ensure that the disk partitions are same on the source and target computers.
For example, If the source computer has a C: and a D: partition, and you copy the cdm
probe configuration to a 
probe on a computer with only a C: partition, the 
probe on this computer will try to monitor a D: partition (which is missing) and generate an error alarm.
Follow these steps:
  1. Log on to the robot where your configured cdm probe resides.
  2. Select the cdm probe from the probe list in the Infrastructure Manager and drag and drop to the
  3. Click 
     from the context menu and enter a unique package name for the copy of the cdm probe archive. For example, rename the package to cdm_master.
  4. Select the 
    Configuration Only
     check box.
  5. Drag and drop this package on any other robot where the cdm probe is already running.
The probe configuration is completed after the probe distribution process is finished.
Edit the Configuration File Using Raw Configure
The raw configuration interface allows you to edit the configuration file or edit the data file.
For example, the following keys can be configured from Raw Configuration:
  • (cdm 6.42 and later) 
    Add this key manually and set the value to
    to prevent clearing of disk alarms on server restarts.
    Default: false
  • (cdm 6.42 and later) 
     Edit this key and set the value to 
     to prevent clearing of disk alarms on soft restarts.
    Default: false
  • (cdm 6.00 and later) get_swap_s: 
    set this key as 
     to use swap -s information on Solaris Global Zone systems.Default: swap -l
    The probe swap -s values can be inconsistent for
    Total Swap Memory + Free Swap Memory
    as the total also includes partial physical memory.
  • (cdm 6.00 and later) nicListCmdPerInterval:
     set this key as
    to generate the SAR (system activity report) using the current value of the network inbound, outbound, and aggregate traffic monitors.Default:
    -  the probe uses the average values in the SAR
  • ignore_device
    : ignores specified disks from monitoring. Navigate to the <disk> section and edit this key. For example, the following key values will ignore all automounted disks that are recognized on each “disk interval”.
    • ignore_device = /autofsmount.*|.*:\/.*/
    • ignore_device = /^(\/proc|[A-Za-z0-9\-]+\:\/.*)/
  • ignore_filesystem
    : ignores specified file systems from monitoring. Navigate to the <file> section and edit this key:
    • ignore_filesystem = /<regular expression>/
The value must be a regular expression that matches all disks or file systems or both. For example, to ignore Auto-mounted disks that are recognized on each "disk interval", use the following:
ignore_device = /autofosmount.*|.*:V.*/
Ensure that you add these two keys manually and then set up the respective configuration.
  • allow_remote_disk_info
    : makes the Device Id of shared drive and local drive identical.
    Default: Yes 
    When the value of this key is
    , the probe generates the Device Id of the shared drive based on remote device. Navigate to the 
     folder and set this key as 
     to enable this feature. This feature is introduced because of the following two reasons: 
    • When file system is mounted on Linux through cifs and the cdm probe is deployed fresh, the Device Id and Metric Id for QoS and alarms for the respective mounted file system are missing.
    • On restarting, the probe is unable to mark cifs drives as network drive and hence generates wrong Device Id and Metric Id.
  • qos_disk_total_size
    : indicates the total size of the disk. Navigate to 
    disk > fixed_default
     and set this key to 
     to enable this feature.
    Default: No
    This key has been introduced in 
    Fixed default
     section under
     for Windows, Linux and AIX platforms.
  • sliced_cpu_interval
    : avoids the delay caused by the probe while executing commands (such as SAR) that cause internal alarms. Navigate to 
     folder and specify a value (in seconds).
  • sliced_memory_interval
    : avoids the delay caused by the probe for collecting data from commands (such as VMSTAT) that cause internal alarms. Navigate to 
     folder and specify a value (in seconds). For example, configure the interval from 
     tab of 
     as 5 min. If you configure the value of 
     key as 10 seconds, then the probe collects data through VMSTAT command for 290 seconds (that is, interval would be (5 min) – 10 seconds). Thus, the probe does not generate internal alarms and seamlessly generates QoS. If the problem still persists, you can increase the sliced_memory_interval by few more seconds. The sliced_cpu_interval and sliced_memory_interval keys have been introduced for AIX platform in the Raw Configuration section. 
  • mem_buffer_used:
    The probe version 5.40 and later supports the physical memory calculation by excluding the buffer and cache memory. To exclude, set the value of 
     key as
    (Default). When the value of this key is 
    , the probe executes the 
    command. As 
    is a CPU intensive command, with a short memory monitoring interval, the mdb command will be invoked within a very short time resulting in increased CPU utilization. CA recommends you to keep the memory monitoring interval greater than or equal to 5 minutes, to minimize your CPU utilization. To include the buffer and cache memory for physical memory calculation, set the value of 
     key to 
Use Regular Expressions
The cdm probe uses regular expressions several times during probe configuration. For example, specifying the regular expression C:\\ in the
Ignore filesystems
field excludes the Disk C of the system from monitoring.
: On UNIX platforms, use the regular expression "/\" to exclude the root directory (/) from monitoring.
A regular expression (regex for short) is a special text string for describing a search pattern. Constructing regular expression and pattern matching requires meta characters. The probe supports Perl Compatible Regular Expression (PCRE) which are enclosed within forward slash
. For example, the expression
matches any character in the range
in the target string.
You can also use simple text with some wild card operators for matching the target string. For example,
expression matches the text
in target string.
The following table describes some examples of regex and pattern matching.
Regular expression
Type of regular expression
Standard (PCRE)
Matches any uppercase alpha character.
Matches with the Uppercase character type of the local disk available on the respective box
Standard (PCRE)
Matches against any character
Matches for all the drives which ends with \\
Standard (PCRE)
Matches against zero or more occurrences of the previous character or expression
Matches for the filesystem name which starts from letter d
cdm FAQs
How to reinitialize disk discovery for multiple robots? For example, I added a new disk on the system and want to rediscover.
To reinitialize the disk discovery, create a custom .cfx file within a configuration package. Use a 'clear' tag within the <fixed> section. This action removes the existing disks, and forces a rediscovery after the next restart.
How to mount CIFS share for monitoring.
The folder to be mounted on Linux should have the "sharing"  option. Use the following  command to mount a windows folder on Linux:
mount -t cifs -o username=administrator,password=<password>//<machine_name_to_mount>/data/windows
Once the drive is added, the type appears as “Network”