Service Engine Status Check (Service Check)

Service Check Category: Netreo Checks

Passive: No

Description

This service check is automatically added to a newly created service engine by Netreo's internal logic (that is, not through a device template or auto-configuration rules) and is generally not intended to be configured manually by a user.

See also Netreo Service Engine (Device Type).

This service check monitors multiple aspects of a service engine. They are listed below and explained further down.

• Minimum resource sizing requirements
• Essential running processes
• Database corruption

These aspects are checked in the order listed according to the schedule configured for the service check itself. If any are found to be failed, the check returns as failed and does not continue checking the other aspects. This first failure becomes the subject of the alarm.

Minimum resource sizing requirements

If the resources available to a service engine appliance are inadequate, the service engine might not function properly. These minimum requirements are based on the number of devices being monitored by a given service engine.

• 500 or greater devices - 8 CPU cores and 16GB RAM minimum
• 1000 or greater devices - 16 CPU cores and 32GB RAM minimum
• 5000 or greater devices - 32 CPU cores and 32GB RAM minimum

If the resource minimums for a service engine under the given load are unmet, the check will return as failed.

Essential running processes

There are several processes that must be active on a service engine for it to operate properly. The required processes are different depending on which types of services are running on the service engine. (See Service Engine for more information about what services are available to be run on a service engine.)

• For the remote collector/remote poller service
  • pollmaster
  • oam
  • mysql
• For the traffic collector service
  • nf_worker
  • nf_listen
  • nf_result
  • nf_cache
  • mysql
• For the log collector service
  • snmptr_syslog
  • syslog_listen
  • eventlog_worker.pl

If any of the processes required for the services assigned to a service engine are not running, the check will return as failed. (Service engines running more than one service require all applicable processes for all running services to be active.)

In addition to the running processes check, a TCP/443 port status check is also made to ensure that the required ports for service engine communication are open and available.

Database corruption

If any tables in the service engine's local databases are found to be corrupt, the check will return as failed and will not continue to look for additional corrupt tables. The name of the corrupt table will be indicated in the alarm.

Check-specific Fields

(See Service Check for configuration parameters common to all service checks.)

• DESCRIPTION
  (Required) This field specifies a name for this check. The name entered must be unique among service check names on the host it is added to (the name may used again only on a different host). It is used to identify this specific check from among other service checks added to the same host.