Re^3: Email Thresholding

Replies are listed 'Best First'.
Re^4: Email Thresholding by bitingduck (Deacon) on Apr 02, 2015 at 15:36 UTC
For something like that I might even add a feature where the on-call person who does the fix runs a little script that sets a flag to indicate the last time a human responded with a fix, and if a new event comes in after that time, but in less than an hour, it sends the email because either the fix didn't fix it or a new event has showed up to set things off and you want someone to respond right away.	[reply]
Re^5: Email Thresholding by mr_mischief (Monsignor) on Apr 02, 2015 at 16:37 UTC
That's a good technical solution but probably a poor choice for the psychology of the situation. You want your hungover, just awakened sysadmin to spend what focus she has available on fixing the actual problem that triggered the alert.	[reply]
Re^4: Email Thresholding by Lotus1 (Vicar) on Apr 03, 2015 at 23:51 UTC
If the process runs every couple of minutes how can you receive over 23000 emails in a single night? It must be sending one email per event. If you sum up all the events each time the script runs and send one email every couple of minutes the most emails you'll get overnight is `15 hours/night * 30 emails/hour = 450 emails/night`.	[reply] [d/l]
Re^5: Email Thresholding by bfdi533 (Friar) on Apr 09, 2015 at 16:46 UTC
Yes, currently the script is sending one email per event. Thus, this thread to get some ideas on thresholding.	[reply]
Re^6: Email Thresholding by Lotus1 (Vicar) on Apr 10, 2015 at 01:43 UTC
Your original post said: When new "events" are seen, an email is sent. ... "An email is sent" implies one email per run of the script. Sending one summary email per run would likely remove the need to threshold or at least give you breathing room while you work on the problem.	[reply]