...
Since the contact information is free form, after the contact information has been obtained it will be necessary to parse it to extract the valid email address(es). The The module to do this should extract the userid before the @ symbol and the host name that follows the @ symbol. These should be validated to ensure the ID and name are correct. I believe the following regular expression can be used to check the hostname:
| Code Block |
|---|
$hostname=~/(([a-z0-9]+|([a-z0-9]+[-]+[a-z0-9]+))[.])+/
|
Or look at http://regexlib.com/DisplayPatterns.aspx you may be able to find a more complete solution
The modification would be needed in checkdata_gif.pl. It may be simpler in many cases to clean up the formatting of the email addresses in NODEDETAILS.
Second step
Once this is done,the next step would be to automate testing so that if we have not been able to gather data from the site for the last 7 days (checkdata_gif.pl provides the number of adjacent days we have been unable to gather data from the site) or more then see if:.
- the monitor is pingable from the host running the script,
- whether the monitor can ping your host (traceroute.pl?function=ping), assuming it has a remote traceroute server
- what a traceroute shows
- the ping_data.pl form appears, also note version of ping_data.pl
If any of steps 1 thru 4 fail then no need to go to following steps you have already limited the scope of the problem.
Then craft an email with the information to review and send to the contacts. This email should identify what is working and where the failure is occurring. It should also point to the FAQ. The email needs to come from a real person in case it gets rejected since the email To: address no longer works.
Eventually if automated the script may run on the 10th, 20th, 28th of the month. These dates are chosen to ensure there are already sufficient days in the month to be able to see enough days with no data.
Requirements
To make the pings one can look at the existing checkdata.pl since it provides the URL to make the calls to ping and can be extended to do a traceroute. However it is probably better to make the pings from the host running the script since it reduces the number of points of failure.