USGS Home Contact USGS Search USGS

monitor_processes

Name:
monitor_processes - An admin utility to automatically report if various time-series processes are not running.

This script was retired from active support in July 2008. This page is available for historical reference and to allow access to the code in case someone wants to try his or her hand in updating it. Use of the script without extensive testing is not recommended and the consequences are your own. If you feel this script was retired prematurely and would like to see it restored send email to GS-W OSW Scripts@usgs.gov.

Information:

This script checks to see whether various time-series processes are running and reports via an email to select users if any are down. It is intended to be run via cron on an automatic schedule of your choosing. Currently the processes monitored are:

Satin
Sentry
Sencomp
NWISWeb transfer
NWISRT replicator

Also, starting with verion 1.1 there is a check of excess files to process via sentry in order to catch issues that are not related to a process failing, such as the LRGS resending several weeks of data.

Email is sent to one set of users whenever one of the processes is found down. Email is sent to another set of users whenever the montor_process job does not run due to ingres maintenance. No email is sent if all processes are found to be running.

Installation:

• Download and unzip the file to your NWIS server to the location of your choice
• Make the script executable (chmod +x monitor_processes)
• Modify lines 16 and 20 of the script to change the "USERID" fields to the email addresses of the selected users. The user(s) in line 16 receive email when the monitor_process does not run because ingres maintenance is running. The user(s) in line 20 receive email when monitor_process finds one of the processes down.
• Set up the script to run by cron on the schedule of your choice.

Example crontab entry:

# monitor satin, sentry, sencomp, nwtransfer and replicator processes (and count of stddata dcascr* files)
15     0-23   *   *   *   /usr/opt/nwis/local/util/monitor_processes > /dev/null 2>&1

Script Language:
ksh

Requirements:

• Nothing in particular

Original Author(s): Joel Johnson, California WSC

Note the original author is noted on this page for the purposes of giving credit only. Please address questions/comments to OSW (GS-W OSW Scripts@usgs.gov).

Known Problems/Limitations/Need Improvements:

Scripts:

Version 1.2: (posted Setp. 11, 2006)

- fixed but that would cause failure to find down processes if run under certain user IDs (like "nwis")

monitor_processes download (Zip file = 2.0K)