Clarification of how sites should remove/add services in the sites
Remove a Service
When a site administrator removes any service from the site, the admins should take the following actions:
Set node production status to non production
Set monitoring off for the node
- Wait for the next scheduled SAM tests and check your node is not in there anymore
- Remove the ldap resource from site bdii
- Deactivate the node
If you miss to do that, NAGIOS tests will continue to be launched against that service, and alarms will be opened in ROD Dashboard.
Add a service
To include a service in the site topology, the site administrator must:
- Add the service in GOCDB (not production and monitored off);
- Add the ldap resource into site bdii;
Set node production status to production in GOCDB.
Set monitoring on for the node in GOCDB.
- Put the node in downtime to check if passes the following SAM test.
Test a service
To test a recently added service with NAGIOS, the site administrator must follow similar procedures as the ones referred in the Add a service section:
- Add the service in GOCDB (not production and monitored off);
- Add the ldap resource into site bdii;
Set monitoring on for the node
- Put the node in downtime to check if passes the following SAM test
Site administrators normally flag the site as "not in production" in GOCDB. However, this does not mean that tickets will not be open against the node. That flag is completly irrelevant since once you tag the node to be monitored, it will generate alarms in ROD dashboard (even if declared as "not in production" in GOCDB, and ROD has to take care of those alarms as usual (open tickets, etc...). To avoid those actions, the best way to proceed is to put the service in downtime.
Index
No Pages to show