IT administrators are spoilt for choice when shopping for network monitoring solutions. We have many vendors flaunting feature sets, pricing sheets, supported vendor list, comparison sheets etc. on their websites and collaterals. While all these are good developments for the seasoned IT administrators, the overkill on technical specifications and marketing ballyhoo can leave the not-so-experienced folks in utter confusion.
Say you are entrusted with finding a network monitoring solution to replace the many free tools you have been using to manage your small network. How do you go about choosing a solution? A good start is to know what the must-haves are and to ensure you have it all covered no matter what you finally end up choosing.
To begin monitoring and managing your network, you will need to import into the monitoring solution the details of the network devices and servers to be monitored. The network may look small now but it will grow and continue to include new devices from multiple vendors, so ensure the network monitoring tool has a vendor-agnostic multi-protocol discovery process, can auto-discover devices and also has provisions to perform a ‘forced’ discovery of a new device/ imported list of devices.
Smart classification & mapping
Different types of network elements – servers, routers, switches, firewalls etc., will have different parameters that need to be monitored for health and performance. The ability of a network monitoring solution to smartly categorize the devices according to their type and vendor will save one a lot of time during set-up. Another important requirement is being able to make custom maps by logically grouping devices to represent clustered environments or geographically distributed resources.
In-depth performance monitoring
One could easily get lost deciding on what to monitor to ensure 100% availability and top performance of the network. If you have been using many free tools, a good start is ensuring the network monitoring solution covers all that your free tools provided. Availability and basic performance parameters can be monitored using SNMP for most of IT infrastructure, of majority vendors. If your infrastructure consists of a mixture of Windows and Linux-based servers, WMI/ CLI (Telnet/ SSH connections) based monitoring capabilities provide an easy means of in-depth performance monitoring for these devices. Ability to monitor traffic flows is also an essential that brings you insights into how your network bandwidth is being consumed, a must for troubleshooting network slowness.
While security for the company’s network and systems is a critical requirement and deserves the care and attention of another full-fledged team, it is important that as an IT admin, one at least knows when fraudulent access attempts are made. By monitoring system log messages that include Windows Event Logs, Syslogs on Unix-based devices, Firewall logs etc., you get to easily know of failed log-ins, failed attempts to secure files, account lock-outs etc.. An added advantage is if the network monitoring solution can extend its security monitoring capabilities by way of plug-ins or third party utilities.
While investigating on alerting features, a must-have is triggering of emails attached with relevant information that can help you quickly correct network glitches. Other value features include triggering and running remediation scripts or programs, provisions for escalation of alarms and setting of device dependencies to prevent those avalanches of alerts. If your company appreciates workforce mobility then you will also want to check out if the network monitoring tool can alert via SMS and be accessed via smartphones. Latest along these lines is OpManager, a network monitoring software that even allows managing of alarms via Twitter Direct Messages.
Scalability & Reliability
Today you could have 50 servers and the very next day you could be managing 100 servers on the company adding a new branch. For the buck spent, it is wise to ensure you get a solution that can handle 5x the load it should presently take care of. Other questions if your company is very serious on having round-the-clock monitoring: Can the monitoring solution alert if the server on which it is hosted fails? Can a redundant failover server take up monitoring tasks and revert when the main server is back up?