Let me first explain what we're trying to achieve with a practical example 
to see if there is a way to achieve it without creating multiple streams 
with identical rules.

Example Stream:
Let's keep it simple
Messages that include Error 500

Alert
Triggered for every message that shows up in this stream
Trigger alert when there are moreless
than  messages in the last  minute and 
then wait at least  minutes until triggering a new alert. (grace period)
When sending an alert, include the last  messages of the stream evaluated 
for this alert condition.

Callback
Graylog Jira intergration plugin
Create a new jira ticket if one for the same endpoint doesn't exist. (ie no 
repeat tickets for the same endpoint)

This all works out fine.

Now we want to add another alert condition so that if we get > 5 errors per 
minute (abnormal situation), we want to send out an email or sms alert.

Easy to create the alert, but how do you create the email / sms callback to 
associate to this alert and not to every other alert.

Basically decouple the alerts from the callbacks and map which alert or 
combination of alerts trigger which callback.

Of course we can achieve this by creating another stream with the exact 
same stream rules and set the alert / callback in this new stream.
But that has a cost, the actual rules match for the stream is more complex 
than a simple one field match, and it's costly to repeat that if we want 
different alert / callback combinations.

How do others do this?
Perhaps using pipelines?
Perhaps we're not fulling understanding stream | stream rules | alerts | 
callbacks flow?

Thanks


-- 
You received this message because you are subscribed to the Google Groups 
"Graylog Users" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to [email protected].
To view this discussion on the web visit 
https://groups.google.com/d/msgid/graylog2/98cec44a-7179-47c2-a921-8dedec785db7%40googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

Reply via email to