User’s Guide – version 1.6 | NetWatch |
Test Community: This performs a simple ‘SNMP Get’ to ensure the correct SNMP community string has been configured for this particular device.
Management Policy: This allows you to decide what tasks will be carried out by each service created by the agent. You can choose which services are managed and whether utilisation statistics are gathered. You can easily change these settings for each service at a later date. The options are:
Use default management settings for new services: The default, this allows NetWatch to decide how to deal with each newly created service. If this is selected most services will be managed when created, but some (for example, ISDN, virtual and dialler interfaces) will be unmanaged. If NetWatch does not have specific support for the type of device you are setting up then all services will be managed when created. If this option is selected no newly created service will collect utilisation information until you choose to collect it.
Manage all new services, but no bandwidth collection: All newly created services will be managed, but none will collect utilisation statistics. You can choose which interfaces to collect utilisation statistics for later.
Manage all new services, with bandwidth collection: All newly created services are managed and those that offer utilisation information will collect it. Note that choosing this option could cause your database to become enormous as utilisation information takes up a lot of space.
Don't manage new services: No new service is managed.
The TCP Port Status service
The TCP Port Status service type can be added to any device. On discovery it scans the device for listening TCP ports within the default “Common Ports” set and creates a service for each one found. A web server, for example, will have port 80 open for HTTP. Once services are created they are polled regularly to ensure they are still active. The options for the TCP Port Status Service type are quite straightforward. The ‘Port Set’ and the contents of a particular ‘Port Set’ that is scanned are configurable. You can also configure a ‘Range Port Scan’, which scans all ports within a predefined range.
Average Response Time: As with the SNMP Interface Test, this is the number of milliseconds a successful check of a port should take. The default is 60 milliseconds since TCP connections can be made or refused very quickly. If you find that you receive a lot of slow response time alerts you may consider raising this value.
Maximum Response Time: As with the SNMP Interface Test, this is the longest NetWatch will wait before giving up on a check and either retrying or reporting the port as timed out. Note that timed out is different to unavailable; if a device is working but the server listening to a port has crashed the device should refuse any request to open a connection to the port rather than ignore it.
37
