Forum Discussion

Keith_Madaras's avatar
11 years ago

Allow for escalation chain stages to start another chain.

Sorry, subject is a bit misleading, its more allow for a rule to hit multiple chains.

7 Replies

  • Below is a specific use-case, yet this could be applied in many situations.

    I have a set of database admins who want to be alerted when any event happens on the production database cluster, yet they shouldn\'t be alerted on events that don\'t relate to the database servers. I also have a critical alert chain that is getting rather complex and should be a catch-all to alert my server team for all critical events. With the current manner that alerting is implemented I would have to add all of the logic of my catch-all critical chain to the database team\'s rule as the service only follows one rule and then stops. This is going to quickly become too complex and rather unmanageable.

    My idea is that alert rules would allow for a manner in which to hit multiple alert chains, therefore if a rule is hit will have the ability to follow multiple chains of action to be sure that I am hitting both the smaller subset of database admins as well as my team who generally see all alerts from all systems.

  • The ability for a single alert to flow through a rule is coming soon... (rules will be able to specify whether to allow alerts that match them to pass through for further processing by other rules, or to stop when they match that rule.)

  • Can we get an update on when this can be added to a sprint? This is really a must have feature that all your customers would greatly appreciate.

  • Anonymous's avatar
    Anonymous

    It's likely this can be accomplished today given the current logic. Obviously, it depends on your needs, but this is very close to an example we give all the time. If you order your rules such that the rule catching the database alerts is first, then the server rule is next, that will properly sort the alerts. If you want the server team also notified about the db issues, they would need to be added to the db escalation chain.

    Perhaps we can help figure out what your rules need to look like if we know more about your needs.