Be the first to know when there's trouble on your network
Have you ever received a call from a customer or, worse yet, your boss asking why a particular service was slow or completely down? More likely than not, you've been in that uncomfortable situation, so you know it isn't fun. A network management tool can help you avoid this situation and ensure that you're the first to know when there's trouble in the network. I recently reviewed four network management solutions, two of which are free, as Table 1 shows. Depending on your environment and budget, one of these solutions is probably right for you.
Spiceworks stands out in this review because it's free. By showing non-intrusive advertisements within the application, you're able to use this tool for no charge. If the advertisements start to bother you or if your company doesn't allow them, Spiceworks can be purchased for $45 a month. You get a small break ($495 instead of $540) if you purchase it a year at a time. This product also stands out from the others because it includes a full-featured Help desk.
Spiceworks is big on community involvement. For example, you can contact local Spiceworks users through SpiceCorps or participate in traditional forums on the community website (community.spiceworks.com). When I browsed through the forums, I found them to be very active and friendly.
Spiceworks is also big on training. For example, you can attend Spiceworks University regional events to get hands-on training from certified trainers, watch the free training videos on Spiceworks TV, and participate in the free IT Webinars.
Installing Spiceworks is easy. As soon as the setup is complete, you create an initial account and password. Then, a large prompt asks you where you would like to start, with the choices being Inventory, Help Desk, or Configuration Backup.
Clicking Inventory starts an IP scan of your network. A wizard helps you set up the proper credentials for Windows, UNIX, Apple, and other servers accessible through Secure Shell (SSH), and for printers, switches, or other SNMP devices. The scan took a few minutes to find, log on to, and inventory all the network devices on my test network.
As soon as the scan was complete, Spiceworks sent me an email detailing what was scanned. The main Inventory page has a running log of what was discovered and notes whether there were any problems with the scan. For example, Spiceworks found my VMware ESXi server but was unable to provide detailed information about it because the username and password that I entered during setup weren't correct. Clicking the ESXi server brought up a menu that allowed me to fix the authentication problem. The next time I ran the scan, Spiceworks found the ESXi server, logged on, and updated the inventory information. Spiceworks not only provided information about each virtual machine (VM) but also listed the names of the VMware data stores and how much space was left in each one.
I have set up and tested quite a few inventory scanners, and Spiceworks got this one right. The interface is very intuitive, making it easy to identify and fix common configuration mistakes.
I spent some time looking over the dashboard, which Figure 1 shows. Out of the box, the dashboard shows the network at a glance and displays information about 12 common areas, such as antivirus software status, Microsoft Exchange Server data, inventory summary, upcoming warranty expirations, and alerts. The dashboard is completely configurable (e.g., you can add, remove, or move sections).
I also tested the Spiceworks Help desk. As the manager of technical support services for a large university, I was curious to see how this solution stood up to the big boys. To help you get started, there are four tickets already created that walk you through an overview of the Help desk system and how it's configured. I quickly learned that the Help desk is very basic. For example, there are no group functions, request types, or escalation paths -- all essential in large IT organizations.
One of the Help desk's strengths is email ticket scraping. This feature is easy to set up and lets users simply email their requests. The Help desk system will "scrape" the information from the email and automatically create a ticket for you. Another homerun is the Help desk portal for users. In addition to emailing requests, users can log on to a web page (which uses Active Directory -- AD -- authentication) to log their tickets.
For a business with just a few people, this Help desk is the perfect solution. Just keep in mind that it won't scale well if you have a large user or technician base.
Spiceworks includes 21 canned reports, including reports that provide Help desk ticket information, list computers without antivirus software, and list computers with low disk space. You can also download community-written reports from the Community website. Two reports caught my eye: One report provides Exchange usage information (e.g., last user logon, users' mailbox sizes, and total number of items) and the other lists locally connected printers.
The Spiceworks mobile application rounds out this feature-rich platform. You can check on Help desk tickets, keep abreast of inventory alerts, check the latest comments from the Spiceworks community, and more. If you support multiple sites, you can set up a profile for each site so that you don't have to remember different username/password/address combinations.
Spiceworks is a great tool. It's fully featured and free. Even at $45 a month or $495 a year, it's a steal. Small companies that don't need or can't afford a big, expensive solution should put Spiceworks at the top of their list.
Foglight Network Management System (NMS)
Sticking with the "free" theme, I chose Quest Software's Foglight Network Management System (NMS) as the second product for this review. You can use this feature-rich software to monitor up to 100 devices on your network for free. If you need to monitor more than 100 devices, you can purchase the full version of Foglight NMS for $99 per device for any device beyond 100. So, for example, if your network has 200 devices, you would pay $9,900 (100 ´ $99), and you can use the full version's three add-ons (Traffic Analysis, IPSLA-VoIP, and Remote Site Monitoring with Pollers) with all 200 devices. You can also purchase two other add-ons: Performance Monitoring and Configuration Management. Unfortunately, the information about these five add-ons isn't easy accessible from the Quest website. If they interest you, you'll have to contact Quest directly.
Both the free and full versions of Foglight NMS include the Network Flow Analyzer Module, Remote Agent Module, VoIP Monitoring Module, and Wireless Monitoring Module. You can also purchase additional components, such as vFoglight for virtual machines (VMs).
Like Spiceworks, Quest is heavy on community. There's a Community link within the product that takes you directly to a community website (www.quest.com/communities) where you can ask your peers questions and even forward ideas to the developer team. However, most of the forums are specifically centered on Quest products and the forums on general topics aren't too active. For example, the SharePoint forum was empty and the Oracle forum only had nine posts, with a majority of them being over 6 months old. Not all of the forums had the same look and feel either.
Foglight NMS should be installed on a separate physical machine or VM. The fully automated setup installs the prerequisites, such as Microsoft .NET Framework 4 and SQL Server Compact. According to the system requirements documentation, SQL Server Compact is for use when you're trying out Foglight NMS or loading device count installations. You need to install the Standard or Enterprise Edition of SQL Server in production environments.
After the setup was complete, I logged on to Foglight NMS Studio and proceeded to register the product. From here, you choose either the free version (up to 100 devices) or enter a license key if you need to monitor more than 100 devices. Even though I was provided a license key, I reviewed the free version.
A splash screen provides links to Help articles that guide you through the initial setup of the monitoring system. You can monitor devices by using SNMP, using Windows Management Instrumentation (WMI), deploying remote agents, or collecting syslog, NetFlow, or SNMP trap data.
I decided to use SNMP and WMI to monitor the devices on my test network. I clicked Add Device(s), which brings up a wizard that lets you add devices through SNMP network discovery or agent deployment. SNMP network discovery did a great job of finding all the devices on my test network. After the devices were discovered, the left-hand menu indicated that 11 devices had the condition of Credential Not Set. Clicking this menu item brought up a list of devices that Foglight NMS wasn't able to log on to. They included an Apple server, network hardware, and Windows Home Server. From this list, I was able to set the proper credentials for each device. The interface was so intuitive that I didn't have to use the Help system or look online for assistance. After I set the proper credentials, Foglight NMS actively monitored my entire test network.
Even though it's a free product, a wealth of features are available. One feature I explored was the network map creation tool. After you add icons that represent your devices, you use a GUI (powered by Adobe Flash Player) to manipulate them into a configuration that simulates the physical layout of your environment. You can create more than one map to show multiple layers. For example, one map can be built on top of a picture of the United States. Each subsequent map can then show more and more detail.
As I worked with the product, I noticed that Foglight NMS didn't properly identify my ESXi server. This could be because it's an older version (3.x). Plus, ESXi was renamed vSphere Hypervisor in 2010. Manually configuring Foglight NMS to recognize the server as a VMware device was easy. After I did this and set the correct credentials, it immediately started gathering statistics on CPU utilization, memory usage, and more. It even included statistics for the VMs.
If your network has multiple physical sites, you might want to have a Foglight NMS server in each location. In this scenario, you can setup different sites to help organize the devices that you're monitoring. These sites then report back to one central Foglight NMS server.
You can set up alerts in Foglight NMS. You can configure it to not only notify you when an alert is triggered but also perform an action, such as stopping a service or running a custom script.
Foglight has the capability to collect NetFlow statistics. Clicking NetFlow in the Tools menu for the first time prompts you to download the 10MB PacketTrap Tool Suite. The suite is a trial version, but it expires in 2031, so you have some time to kick the tires before it runs out. (According to Kelly O'Dwyer-Manuel, a senior analyst/public relations specialist at Quest, this tool will be rebranded to Quest Free Network Tools in the near future and the trial version will simply be turned into a free application.) However, I wasn't impressed with this "bolt-on" approach. Foglight NMS by itself is an impressive application, but the way PacketTrap integrated into the product wasn't clean.
Foglight NMS is a well laid out application with a great dashboard, as Figure 2 shows. If you have a Help desk that needs to view the status of your network, you can provide a read-only view via a website. You just add the port 5053 to the end of the server name (e.g., https://Foglight:5053), and you're set.
Overall, I found Foglight NMS to be intuitive and easy to use. The fact that it is free for up to 100 devices should make mid-sized businesses perk up and take notice. I've been looking for a free network monitor for my home network -- and I think I just found it.
Foglight Network Management System (NMS)
Ipswitch's WhatsUp Gold is a mature product that's been around for well over 10 years, which is evident in its deeply rich feature set. It's obvious that Ipswitch has been listening to users and making improvements based on their comments and suggestions.
WhatsUp Gold comes in three editions: Standard, Premium, and Distributed. The Premium Edition includes everything in the Standard Edition and adds monitoring through WMI, UNIX and Linux monitoring, Wireless Access Point monitoring, and other advanced monitoring features. The Distributed edition adds functionality for networks that are distributed over a large geographical area. You can find a detailed list of what comes with each edition on the WhatsUp Gold website. If you find that you need to move up to a higher edition, no reinstallation is necessary. You simply enter a new license key, and the product is automatically upgraded to the new version
I reviewed the Premium Edition of WhatsUp Gold v15. Setting it up was easy because it takes care of the prerequisites, such as installing .NET Framework 4.0, IIS, and SQL Server Express 2005. If you have a large network, you might want to use a dedicated SQL Server 2008 or SQL Server 2005 instance instead of SQL Server Express. The setup package contains two WhatsUp Gold add-ons: WhatsConnected and WhatsConfigured. For the purpose of this review, I installed only WhatsUp Gold.
After the installation, a wizard walks you through configuring SMTP notification and credentials for the different devices connected to your network. As the network is scanned, you can watch WhatsUp Gold find each device, resolve its hostname, and attempt to discover its role (e.g., VMware host, Windows server).
WhatsUp Gold did a good job of finding and identifying the devices. What it couldn't properly identify was easy to adjust. What's nice about this product is that you can configure everything using either the Windows console or web console.
Larger networks can have hundreds and even thousands of network devices. WhatsUp Gold helps you find these devices by grouping them together. There are two kinds of groups: nondynamic groups and dynamic groups. Nondynamic groups are manually managed. You drag and drop devices onto each group manually. For example, if you have two physical locations, you might want to create a New York group and a Seattle group, then drag and drop the network devices into the appropriate group.
WhatsUp Gold comes preconfigured with 20 dynamic groups, and the Help database has another 14 more. You use a SQL statement to populate them. For example, if you want to place all of your Cisco devices into a dynamic group, the SQL statement would look like this:
WHERE Device.bRemoved = 0 AND
Device.sSnmpOID LIKE N'18.104.22.168.4.1.9%'
Even if you're only a little familiar with SQL, the examples should be enough to help you create your own dynamic groups. However, if you're uncomfortable with SQL, you might find this feature frustrating. I would like to see WhatsUp Gold keep the manual SQL functionality (for those users who know SQL) but add a wizard that helps you build SQL statements.
Ipswitch posts the database schema on its website, which makes writing database queries and creating custom reports easier. Obviously it doesn't support writing to it outside of the WhatsUp Gold application, but the company encourages read-only access if it helps you gather the data you need.
Besides performing the typical ping check, WhatsUp Gold monitors performance (e.g., CPU utilization, memory usage) and services (e.g., DNS, HTTP). It also has what's called "Passive Monitors" for monitoring SNMP traps, syslogs, and Windows event logs.
I found that the Windows console is easier to use when setting up and configuring the monitoring tool, whereas the web console does a better job of displaying the status of each device. When a device goes down, the device's icon turns yellow and the web console displays a message noting which monitor (e.g., ping monitor, DNS monitor) is down and for how long. If the device continues to be unresponsive, the icon turns red and the time down gets updated. Clicking the device brings up additional information to help you troubleshoot the problem.
WhatsUp Gold has a very configurable dashboard. I found it easy to add, remove, and adjust the dashboard to my liking, as Figure 3 shows.
There's no doubt that Ipswitch has been around for the long haul and is well respected in the industry. Its latest version of WhatsUp Gold doesn't disappoint.
WhatsUp Gold v15
Orion Network Performance Monitor (NPM)
SolarWinds' Orion Network Performance Monitor (NPM) is another product that falls into the mature category. Although it's pricey when compared to the free products in this review, you get a lot of bang for your buck.
Like the other products, I setup Orion NPM on a dedicated Windows Server 2008 server that's a member of my test network's domain. Installing it on a domain controller (DC) isn't supported. The setup routine takes care of the prerequisites for you, including installing .NET Framework 3.5 with SP1 and IIS. A backend database is required to store the data that's collected. You can use SQL Server Express 2005 for a test or small network. However, SQL Server 2005 or later is recommended.
The first time you log on to the web-based administrative console, a wizard helps you configure the credentials for SNMP, VMware, and Windows and configure the IP addresses you want to scan. The results are then imported into the database.
Orion NPM found each device on my network without any problems, including the ESXi server. When I expanded the ESXi icon, a complete list of the VMs being hosted was displayed. Hovering over each VM provided a summary of that machine's status. If the VM wasn't running, its name was displayed in gray text.
Instead of installing agents, Orion NPM monitors devices using Internet Control Message Protocol (ICMP -- aka pings), SNMP, WMI, or syslog, or by logging on to the device. According to the Orion NPM Administrator Guide, it uses agentless methods for the following reasons:
- NPM does not install outside agents on your mission-critical servers.
- NPM does not employ services that take vital resources from critical applications.
- NPM does not install any code on monitored network devices. Unmanaged or outdated code can open security holes in your network.
For me, this clearly answers the age-old question of whether agent or agentless monitoring is best.
I found Orion NPM's dashboard easy to navigate and well laid out. For example, as Figure 4 shows, it includes handy tabs at the top, such as Top 10, Alerts, Syslog, and Events. Like the ESXi server and its VMs, each device being monitored can be clicked to bring up a more detailed screen. The detailed screens have multiple "speedometers" that show real-time statistics such as average response time, packet loss, average CPU speed, and memory used. Clicking a speedometer displays a graph of that measurement's history. There is also a customizable historical graph that shows the device's status over several weeks or months.
You can import the Network Atlas tool into Orion NPM. This tool lets you create maps of your network in four easy steps: Choose a background, add objects to the map, connect the objects to the back-end database objects, and customize as needed. SolarWinds provides more than 30 maps, including maps of the world and individual continents. In under a minute, I created a map of the United States, added an icon, and linked the icon on the map to a switch in my test network.
SolarWinds has an interesting licensing model. Instead of a flat per-device price, the total cost is calculated by finding the largest number of interfaces, nodes, or volumes. For example, if you have two 48-port switches, 100 servers/nodes, and 20 hard drive volumes, you only pay for the highest number -- in this example, 100 servers/nodes. When I entered this information into the online license calculator, I came up with a price of $3,595. The rest of the devices that you want to monitor on your network are essentially "free." Before you buy online, though, be sure to give them a call to make sure you don't purchase more licenses than you need.
Quite a few configuration-related tools are available. For example, have you ever set up an alert for a particular device, only to find out that you weren't notified when the device went down because you set up the alert incorrectly? Orion NPM includes the Test Fire Alerts tool, which lets you verify that you'll receive an alert when the device goes down.
One odd aspect of Orion NPM is how its individual tools are implemented. Instead of putting all the tools in one executable, many of the tools are separate programs that you need to install. I counted 15 separate programs.
There's no doubt that Orion NPM is a proven network management program. I found it to be very robust and capable. However, the unique licensing model can be a bit confusing and the price might be too high for some companies.
Orion Network Performance Monitor (NPM)
Each of these products provides great management capability. In looking back on my experience with them, I believe that they each fill a specific role: Spiceworks is great for the smaller business with little or no budget. Spiceworks also works well if you or your company supports the IT infrastructure of multiple smaller organizations.
For mid-sized businesses that need an inexpensive network management tool, Foglight NMS fits the bill perfectly. You can monitor up to 100 devices for free. As time goes on, you can easily add devices and let Foglight NMS grow as your requirements grow.
Large businesses that want the absolute best in network management will want to take a serious look at Orion NPM.
WhatsUp Gold provides the ultimate in network management. Although it's significantly more expensive than the two free products in this review, it also has significantly more features as well. After hours of working with WhatsUp Gold, I was still finding cool and interesting features to try out. If you have a large network, have a lot of devices, and demand the ultimate in network monitoring and management, then WhatsUp Gold is your best bet.