Automate your network monitoring with one of these comprehensive products
At the midsized company I work for, we've been using an older network monitoring system for about seven years. It gives our administrators basic up or down information about servers and services, and it sends SMS (text message) and phone alerts to our cell phones in the event of problems. I decided it was time to upgrade or at least add a more current tool that could provide improved performance and detailed status information about our Exchange Server, SQL Server, and Terminal Server systems on the network at an affordable price. I came up with three viable candidates: Ipswitch's WhatsUp Gold Premium, ManageEngine's OpManager Professional, and SolarWinds' ipMonitor. All three of these network monitors cost less than $3,000 (for 100 devices) and are available for trial periods so that you can test drive them for yourself. Let's see how they stack up.
To prepare for testing, I first needed to enable SNMP on all devices, including our Windows servers. I set up read-only SNMP access by configuring the SNMP settings for each device on the network that I wanted to monitor.
In Windows Server 2003/2000, you install the service from Add/Remove Programs, Windows Components; in Windows Server 2008, you add the SNMP feature from Server Manager. Then, you need to go into the Control Panel Services applet and configure the SNMP service—a quick and easy process. Managed network devices such as firewalls, switches, routers, and printers will also have SNMP management capability and are usually quite easy to configure. For more information about SNMP, see the Microsoft appendix "Simple Network Management Protocol" (technet.microsoft.com/en-us/library/bb726987.aspx).
Next, I installed all three monitoring systems on one of two Windows XP SP3 machines at my desk. Once installed, each system consisted of a database and web server. In all three monitoring systems, the system is meant to be used from the web server interface by multiple individuals, and you can configure credentials with full or varying levels of limited access. Each user to either system has the ability to add, remove, and reposition widgets on his or her dashboard. Widgets give insight to a particular aspect, such as processor or memory utilization, across many devices on the network.
Before using each monitoring system to scan the network (called the discovery process), I added credentials that the monitoring system would use to gain access to each device that it discovered on the network. Ipswitch's WhatsUp Gold Premium has credential options for SNMP, Windows, Telnet, SSH, ADO, and VMware; ManageEngine's OpManager Professional has options for SNMP, Windows, Telnet, and SSH; and SolarWinds' ipMonitor has options for SNMP and Windows.
After configuring SNMP at the network devices and credentials (Windows and SNMP) for each of the three monitoring systems, I ran the discovery process on a range of IP addresses in my local subnet. All three monitoring systems discovered about 70 devices. Using the default scanning options, all three monitoring systems did a good job identifying each device type, as well as providing a lot of insight into device status. All three included several monitors for what you would expect in a network monitoring system, such as processor, memory, disk usage/utilization, packet loss/latency, Exchange/Lotus, Active Directory, and every Windows service on the system. All three products had the ability to add monitors to each device one at a time or large groups of devices all at once.
Both OpManager and WhatsUp Gold include interfaces for identifying and collecting VMware events on VMware host and guest servers. Also, both OpManager and WhatsUp Gold contain a switch port mapper, which shows you what device is connected to each port on your managed switches. With this information, you'll know which port on each switch contains the link to critical business applications without having to physically trace wires in your server rooms. Then, you can configure alerts on the switch devices for these individual ports. With OpManager, it's easy to get the switch port mapper results by selecting the switch and running the Switch Port Mapper tool; the product returns results within a few seconds. In the case of WhatsUp Gold, the tool is called MAC Address and needs to be run with the Get connectivity option turned on. WhatsUp Gold takes much longer to get the results, as it appears to be scanning and calculating connectivity information for the entire subnet.
With both ipMonitor and OpManager, I would occasionally notice strange readings that made me wonder what was going on. With ipMonitor, when processor utilization reached a low percentage, it would sometimes show as a negative value on the dashboard widgets. In another instance of low processor utilization, ipMonitor sent me an alert that a processor was at 11,490 percent utilization! OpManager would sometimes track and send me accurate information about domain controller (DC) disk utilization but not show it in the top 10 disk utilization widgets on the dashboard—but right above that widget was another widget showing me that one of our DCs should be in the top three. No such situations occurred when I used WhatsUp Gold. In fact, WhatsUp Gold tracks processor cores in its processor utilization widgets, and when I compared the WhatsUp Gold processor widget reading with the Windows Performance Monitor results, it always appeared accurate on all individual cores. Likewise, hard disk utilization was accurately reported on all hard disk dashboard widgets.
WhatsUp Gold includes a monitoring library, which lets you create new monitors based on other monitors. Larger organizations might find this feature useful for ensuring a uniform set of monitoring by device type across a large technology environment; this seems the most efficient way to adjust monitors on groups of devices.
WhatsUp Gold doesn't contain vendor-specific monitors (except for an APC UPS monitor)—as OpManager does for Dell, HP, and IBM—but rather contains an option for Active Script monitors, letting you use VBScript or JScript to write your own monitoring routines. There's also an online resource center for Active Script, where the WhatsUp Gold user community can download and share code.
One improvement I'd like to see WhatsUp Gold incorporate would involve the UI—which Figure 1 shows—mainly because it's so linear. For example, after drilling down into the Active Monitor Library, it takes about five Cancel or Close clicks to return to the dashboard. Also, unless scripted, WhatsUp Gold doesn't have a monitor for checking the status of a website, which might be necessary—especially if it's hosted on somebody else's server that you don't have another way of accessing.
For situations in which a device stays down, you can configure WhatsUp Gold to send alerts at 2, 5, or 20 minutes. So, it's possible to escalate the alert in situations when the expected individual doesn't respond within that time frame.
Of the three tools, WhatsUp Gold is the only one with the option to integrate with an LDAP environment, which could be particularly valuable to large deployments.
Upon installing OpManager, I noticed that it was simple to navigate and configure a very wide range of features. Along with SMS and email, OpManager has the option for sending a Direct Message to a Twitter account—a nice alternative to email. Using the Twitter account this way gives me an overview of what's happening on the network, but because my phone doesn't ring when a tweet comes in, I still want the text-message alert for anything critical. I can see myself setting low thresholds for a Twitter alert on all server devices, just to get a good running log of what's going on, but not necessarily using it for warning of critical situations.
In addition to its expected monitors, OpManager offers vendor-specific SNMP performance monitoring for Dell PowerEdge, HP ProLiant, and IBM Blade Center devices. OpManager can also integrate with the Google Maps API so that you can add your devices to a Google map. However, you'll need to pay for a Google Maps API Premier account (unless you plan to make your network map publicly available), as this is a licensing requirement for the free version of Google Maps API.
For situations in which an administrator receives an alert but doesn't respond within a certain amount of time, you can configure OpManager to send another alert to a different administrator. For example, for an individual who might be out sick or distracted but would normally respond to a critical event on a particular group of servers, you could have an escalation alert notify another administrator if the original alert isn't acknowledged or cleared within a specified number of hours/minutes.
OpManager was also the only one of the three products to include a section specifically for monitoring the quality of VoIP over the WAN. The VoIP monitoring tools in OpManager require devices at both the source and destination networks to support Cisco IP SLAs. In addition, OpManager—whose UI you can see in Figure 2—includes more monitors and dashboard widgets than either of the other two products.
When I first installed ipMonitor, I found the UI—which Figure 3 shows—confusing. For some reason, it took me forever to find the setting for how often the product would check individual system monitors (which turned out to be every 300 seconds). However, after using ipMonitor for several weeks, I found it to be extremely easy to use, offering enough monitoring options to keep a good eye on the network. With ipMonitor, you can adjust the default scan so that any service or performance option will always be selected in future scans. In addition to the expected (and previously mentioned) monitors, ipMonitor offers a Windows event log monitor that you can configure to check for events that you want to be alerted about.
On the downside, ipMonitor doesn't contain any type of alert tracking/assignment system. If you have only one network administrator, this should be no big deal, but larger IT departments will probably find that the inability to acknowledge/assign/clear alerts is a problem. Unless the administrators remember to communicate outside the system, multiple administrators could receive the same alert and start working on the same problem. Perhaps devising an agreed-upon way of responding—for example, if you get the alert, it's your responsibility to take care of it, or dividing up the network devices so there's no question about who should respond—would do enough to prevent this situation.
I've just made my decision about which of the three products will best suit my environment. I'm going with ManageEngine OpManager, licensed for 50 devices, for a few reasons. First and foremost, I want to be able to monitor as many aspects of our environment as possible because I realize it's the best way to avoid unexpected outages. OpManager definitely offers the most breadth in this area. The second reason is budget. I’ll continue to use our existing up-or-down monitoring tools for our workstations and printers, so I can avoid paying for the additional licensing. Third, I really like the way the people at ManageEngine have designed OpManager to take advantage of new technology, and I feel it's worth paying the annual service and support fee to get the updates as this product develops.
Ipswitch WhatsUp Gold Premium
Pros: Provides the most accurate information of all three products; lets you create custom monitors; offers comprehensive VMware monitoring tools, integrates with AD
Cons: Not as many built-in monitors as the other products; more expensive than the other two products (until you get up to 500 nodes)
Rating: 4.5 of 5
Price: $7495 for 500 devices, $2,695 for 100 devices; $2,195 for 25 devices
Recommendation: For IT shops that have a large VMware installation or a desire to create their own monitors
Contact: Ipswitch • www.ipswitch.com • 800-793-4825 or 781-676-5700
ManageEngine OpManager Professional
Pros: Best overall UI; more built-in monitors than the other two products; lowest pricing for 50 or fewer nodes
Cons: Didn't monitor all devices accurately in my tests, so it might require some troubleshooting to get it completely functional
Rating: 4.5 of 5
Price: $7194 for 500 devices, $2394 for 100 devices; $1194 for 50 devices; $714 for 25 devices
Recommendation: IT shops looking for the maximum amount of built-in functionality (with the exception of AD integration) will appreciate OpManager Professional. In the 26- to 50-device range, it's about half the price of the other two products.
Contact: ManageEngine • www.manageengine.com • 888-720-9500 or 925-924-9500
SolarWinds ipMonitor
Pros: Unlimited devices at a very low price; easy to use
Cons: No multiple administrator communication mechanism
Rating: 4.0 of 5
Price: $1,995 for unlimited devices (25 monitors free)
Recommendation: If you're on a tight budget and you need to monitor a lot of devices, or if your monitoring needs aren't terribly complex and you're OK using an outside-the-system way of communicating between administrators, SolarWinds is for you.