Introduction to monitoring

What can be monitored?

nVision can monitor the following:

•Host status
This is monitored for every host and it allows you to get reports about each host availability over time.

•Services

oAvailability: in case the service stops responding nVision will present such information on the map and can raise an alert.

oPerformance: response time and percent of packets lost. You can monitor any TCP/UDP service. There is a large predefined list of available services including MS SQL Server, Oracle, Notes/Domino, etc.

•Mail and web servers
Specific service tests: nVision has several built-in probes that can check the performance of several high level features of service. It includes probes such as:

oWeb page load time – measures the time of loading a specific web page.

oWeb page contents change – checks for any change of web page contents.

oPOP3 Login time – measures the time it takes to successfully login to a POP3 server.

oSMTP send time – measures the time it takes to successfully send an e-mail with an SMTP server.

•Routers and switches (MRTG)

oNetwork interfaces: status and in/out network traffic.

oSwitch ports: the information about every port status, MAC and IP of computers connected to any port and their total network in/out traffic.

oHost network traffic: the network traffic generated by a network device (monitoring RMON with SNMP)

•Performance counters

oSNMP: you can monitor any SNMP counter that returns numeric values.

oWindows: nVision allows monitoring Windows counters, which helps to monitor system performance itself and also the performance of applications running on it. So you can monitor counters of such services like MS SQL Server, Exchange server, etc.

Visualization

nVision has the ability to present all the monitored parameters (both for services and counters) on clear charts. Not only you get the reports of the values over time, but you can watch them in real-time also.

Monitoring time

When you setup the monitoring time in the host properties, it does not mean that the services and counters will be monitored exactly every such period. When nVision has to monitor a large network with many hosts, the monitoring interval may increase because nVision can send only a specific number of requests per second. Consequently, host monitoring time is just the shortest time that can be used for monitoring services and counters. In case of several number of hosts this time may increase significantly.

How monitoring data is handled

nVision collects all monitoring data in the program’s memory first. This information is gathered in the form of consecutive probes stored at the time of each poll. You can see all of those probes on the 15-minute chart only. After all gathered data exceeds the limit of allocated memory then the oldest probe is removed every time a new one is added.

Monitoring information is saved to the database in 1-minute averages. Therefore, when viewing performance charts for long intervals you can see those data with a 1-minute resolution at best. nVision does not save all probe values because the program is designed to monitor large networks. In such networks with many hosts, the amount of data gathered every day is significant and it would be not possible to process it in reasonable time. It would also fill up even a very large hard drive very fast.