Router interface checks (LSPs)

This forum was archived to /woltlab and is now in read-only mode.
  • Hi All


    I am moving some router checks from an old Nagios installation to Icinga2.


    I have working checks but was wondering if my config was efficient or if perhaps there is a better way.



    This has the effect of creating many services (one per interface) and this is one of the smallest examples I have.


    I also need to do similar checks for interface errors etc. and active interfaces are obviously different on each router.


    Generating the config is not really an issue as I have scripts that poll the routers for the details.


    Is it bad practice to create so many Services and if so could you point me towards a better way?


    Thanks and regards

    Derick

  • couldn't you make better checks via SNMP and collecting all the data into on service?

    Linux is dead, long live Linux


    Remember to NEVER EVER use git repositories in a productive environment if you CAN NOT control them

  • How does the CheckCommand check_alcatel_lsp and its plugin look like? Maybe they allow you to check multiple interfaces at once (like check_interfacetable_v3t).

  • Thanks for the advice . I did a proof of concept for a new check in Python and I can absolutely do it in one check.


    The only problem is that currently our NOC is able to ACK alerts for individual interfaces going down.

  • for this, you would have to run a discovery on the network device, feed that data into the icinga2 host object and then trigger a service apply that runs on said data.

    Linux is dead, long live Linux


    Remember to NEVER EVER use git repositories in a productive environment if you CAN NOT control them

  • No I meant if an interface is down but shouldn't be our noc can ACK the interface while they investigate the issue. We would not want to remove the check if it is supposed to be up.

  • you missunderstood me. what I meant was the automated service generation so that every single interface has its own check. that way you can ack a single port failure without disabling notifications for the rest of the ports.

    Linux is dead, long live Linux


    Remember to NEVER EVER use git repositories in a productive environment if you CAN NOT control them