Report generation using script

UNIX for Beginners Q & A
1

Hi all

I have a unix script that generates a report with the following information:
uptime, mounted file systems, disk usage (> 90% --> critical, <75%-90%> --> warning, < 75% healthy), Mem usage, CPU usage and load average.

But I would like to create one single report containing all this information but for 10 servers.
Please can you give an hint in how do I go about it

2

Ever considered deploying Nagios?

Addendum:
More seriously, many options you have and we can't decide for you... About Nagios: Depending on your aim (looking at what your script returns...) It could be of some interest...
Q:
1- Many hosts, multi platforms?
2- Any other monitoring tools being used?
3- If only this script, how are you returning/look at the logs of other servers? Can an alternative method be acceptable?
...
...

3

Yes, I did consider nagios, and even setup the host server, but I am struggling to add hosts (clients).
They are (clients) hp-ux, solaris and centos servers that I want to monitor:
its nagios core version 4.0.8.

4

But for now what is needed is something very simple, like an excel spreadsheet with a summary information of the health status of the servers

5

Cannot you simply iterate over the list of hosts calling your existing script for each host spooling the results to one aggregated report?

6

Here is how I might approach this:

  • Install your script on all 10 of your servers and make sure it works for each.
  • Optional: Install a secure (HTTPS) web server on each server.
  • Optional: Configure a page to use PHP or Python (for example) to call your monitoring script.
  • Optional: Test to make sure you can call this script from an HTTPD call. You can secure this script as required.
  • Modify your script to return your monitoring information (either as an API or write to file) as a serialized JSON object (your API).
  • Optional: Test and confirm you can call your API.
  • Optional: Install this API on all servers.
  • Note: If you do not want to install a web server, you can do this with other methods, for example write the JSON object to a file and use rsync .
  • Optional: Create a web page on a secure server which will call all 10 APIs and display the information on a single web page.

Not knowing your environment and other considerations, if I was doing to do this, I would go with the secure web solution, create the API in PHP to call your monitoring script and format the API return (serialized JSON object). Then I would write another app to call this API (in a loop) for each server and display the results in a dashboard with some canned dashboard dials using a standard Javascript lib for this.

Having not seen your script, I estimate this entire project can be coded and tested in less than a day, two days at the most considering there are 10 servers.

Hope this helps.