sawyl: (Default)
An unexpectly productive day after one of my colleagues finished provisioning the new Nagios XI server and passed it over for configuration. After investigating the bulk import feature, I decided it would probably make more sense to recreate the hosts, groups, and services using the database.

Fortunately, thanks to the way the systems — with the exception of the Lustre appliance, which we didn't configure — were templated, it was a relatively simple task get everything defined and within a couple of hours, I had all the TDS monitoring up and running. Which, because the systems are almost indentical, means that all I need to do to duplicate the monitoring on the production systems is to duplicate the host entries. And while the are a lot of production nodes, it shouldn't be too formidable: we can either use the new ability to clone systems or possibly even create a custom wizard to set them up.

The process of mirroring the definitions over from the Lustre appliances — they run Icinga v1 under the covers — is likely to require more effort, but at the same time, it's also less urgent because the storage is already being monitored and because it requires me to re-write the code used to push alerts on to the XI server.
sawyl: (Default)
Via RW, a neat method of using the Nagios configuration files to quickly switch large numbers of systems from one hostgroup to another, e.g. to allow machines to be moved from test to production status.
  • define a host template that conatains a hostgroup directive that puts the host in a primary group, e.g. test or production
  • define a host using the template
  • add any additional groups to the host using hostgroup +<secondary-group>
  • finally, toggle the value of the template as necessary

Although this trick doesn't really apply to the way we do things, moving the assignment of hostgroup memberships from the group definitions to the host definitions greatly simplifies the configuration files and makes it much easier to add and remove systems.


sawyl: (Default)

September 2017

      1 2
3 456 78 9


RSS Atom

Most Popular Tags

Style Credit

Expand Cut Tags

No cut tags
Page generated Sep. 26th, 2017 02:27 pm
Powered by Dreamwidth Studios