Software User's Guide
The SPS/Disk Space Manager (DSM)

Table of contents

  1. Introduction
  2. The sps_dsm.pm program
  3. The Control Table (sps_dsm.table)
    1. logical_name
    2. object_path
    3. size
    4. log_syserr_msg
    5. select_latest
    6. notify_user_1..5
    7. critical_limit
    8. critical_message
    9. critical_command
    10. ...interval
    11. warning...
    12. available...
    13. normal...
  4. The SPS/DSM Log File

Introduction

The SPS/Disk Space Monitor resolves two common and serious problems. The first is the ongoing struggle to control and manage the amount of valuable disk space consumed by each application, user or group of users. The second is the ability to detect quickly and to correct automatically conditions of insufficient free space which may cripple the system, cause critical applications to fail and create production disasters.

DiskManager runs in the background and is started by the VOS start_process command. A simple configuration table controls DiskManager. Given the flexibility of defining and automatically executing any commands when an object size reaches its limit, the system administrator can build and tailor his own DiskManager environment(s). For example command macros and/or programs can broadcast messages to the operator's and user's consoles, stop a process, delete *.out and *.kp files, run cleanup jobs and more. Please note that DiskManager is used for detection and prevention of disk space problems, it does not physically put any limitations on other processes.

The product works with three types of thresholds: Warning, Critical, and Available. Different threshold will allow you to take different action and produce different messages when the space-used reaches different levels. At least one threshold must be specified. Using The Available will produce an alert when space-used goes down BELOW the specified threshold. This means that you will get notified not only when you run out of space, but also when space becomes available.

The sps_dsm.pm program

sps_dsm.pm

Purpose

The sps_dsm.pm command initiates the disk space watchdog program. To start the program, use the VOS start_process command.

CRT Form

---------------------------- SPS/Disk Space Manager --------------------------
 -control:               %softmark#d03>SPS>disk_space_manager>sps_dsm.table
 -ignore_page_partition: yes
 -output_path:           %softmark#d03>SPS>disk_space_manager>logs>sps_dsm.html
 -email:                
 -lams_input_q:          %softmark#d03>SPS>alert_manager>logs>sps_lams_input_q

Lineal Form

sps_dsm.pm  [-control]  [--ignore_page_partition]  [–output_path]  [-email]
            [-lams_input_q]

Arguments

-control

The relative or full path name of the Manager's control file (sps_dsm.table).

-ignore_page_partition

If set, calculation will be done without including the Paging partition (if any) - only the File partition will be use.

-output_path

The relative or full path name of the report file.

-email

When producing HTML-formatted reports, you may choose to send it via your E-Mail Server to selected users. If so, simply enter their email addresses as you would in your favorite email program.

-lams_input_q

The relative or rull path-name of the AlertManager Input queue. This optional interface allows ACM to send aleter message to AlertManager.

The Control Table (sps_dsm.table)

This table contains all the information DiskManager needs to run and successfully perform its function. To create the control table, execute the create_table command.

organization:			relative;
index :				logical_name  no_duplicates;
fields:

logical_name			char (32)  var,      
object_path        		char (256) var,   
size                		dec  (9), 
select_latest			bit (1) aligned, 
log_to_syserr			bit (1) aligned, 
notify_user_1      		char (65)  var,  
notify_user_2      		char (65)  var,   
notify_user_3      		char (65)  var,   
notify_user_4      		char (65)  var,
notify_user_5      		char (65)  var,
user_exit          		char (32) var,
critical_limit     		dec  (9),         
critical_message   		char (80)  var,  
critical_command   		char (256) var,  
critical_interval  		bin (15), 
warning_limit      		dec  (9),         
warning_message    		char (80)  var,  
warning_command    		char (256) var,  
warning_interval   		bin (15), 
available_limit          		dec (9),         
available_message    		char (80)  var,  
available_command   		char (256) var,  
available_interval   		bin (15), 
filler				bin (15),
normal_message    			char (80)  var,  
normal_command    			char (256) var,  
normal_interval   			bin (15), 
end;

logical_name

 A unique name that identifies the monitored object. This name will appear in DiskManager reports and error messages.

object_path

 The path name of the monitored object which could be:

  • A directory path name

  • A file name

  • A link to a directory or a file

  • An entire disk pack name (on current system only)

    You may use star names (e.g. *.out). If you do, the star name must pick at least one file. Start names for directories cannot be used.

    When using DIRECTORIES, you must have at least STATUS access to the specified directory and to all its subdirectories.

size

 The size of the object. What would be considered it's 100% capacity.

log_syserr_msg

 When set (1), the server will write a message to the daily system error log file.

select_latest

 This field is relevant only if a star name is specified in the object_path field. When set (1), only the last-modified file will be looked at, otherwise (0) the size will be calculated based on all matching files.

notify_user_1..5

 Specify up to 5 user name masks to receive messages. Example: "Bob.*" or "*.SysAdmin".

critical_limit

 The Critical threshold as a precentage of "size".

critical_message

 The Critical Level Message. You can make up your own messages and use the following keywords:

  • @obj The object's name.

  • @size The object's current size.

  • @limit The limit specified in the control table.

  • @per The percentage of space used to size limit.

  • @time The time of the event.

Example:

'Critical Problem on @obj S:@size L:@limit ==>>@per'

May produce the following message:

Critical problem on #d38 S:180000 L:170000 P:106%

critical_command

 The Critical Level command to be executed. As in the Message field, you can use the special keywords that start with the @ sign to be substituted with the proper values at runtime.

...interval

 A time in minutes between each checkup while the object is in the Critical state.. A minimum of 1 minutes is allowed.

warning...

 Define the Warning condition.

available...

 Define the Available condition.

normal...

 Define the Available condition. Unlike Warning & Critical Levels, Normal Level "Alerts" (sending messages, executing the command line) will be performed only once, when the level changes from Available/Warning/Critical back to Normal.

The SPS/DSM Log File

The following is as a sample of the daily log file (in the logs sub-directory). 

14:48:01 1. SYSERR-1...........................syserr_log.(date)
14:48:01 2. SYSERR-2...........................syserr_log.(date)
14:48:01 3. My Home Dir BT.....................#d02>>BT
14:48:01 
14:48:01 Status   Logical Name / Path                %    Size   Limit  	Type
14:48:01 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
14:48:02 Avail.   SYSERR-1                          60%      6      10	File
14:48:06 Critical My Home Dir BT                   111%    608     550	Dir.
15:18:03 Avail.   SYSERR-1                          60%      6      10	File