Server Infrastructure Tools, Monitoring Software

by Nelson King

Keeping an eye on your servers requires more than eyeballs. Server monitoring tools oversee server operation, traffic, and usage. We look at what organizations are monitoring and provide a matrix to help determine which products meet your needs.

Monitoring many servers — i.e., keeping an eye on their operation — is one step toward meeting two priorities: troubleshooting and optimizing.

Hardware may fail, and software may not perform well enough. Servers may fail entirely, or worse, appear to be operating while they are no longer performing vital functions. The more servers a company is managing, the greater the likelihood of problems and the more difficult it becomes to monitor them. Then there's the not so small matter of getting bang for the buck. Are the servers performing well enough? Can a group of servers be considered reliable?

These and similar considerations are behind the need for specialized software that falls under the category of server monitoring tools.

Getting the most from server monitoring tools is more than a numbers game of managing as many servers as possible with the fewest number of people. It's also a matter of what is monitored. In most cases, this means three areas:

  1. Monitoring server operation (the running status)
  2. Monitoring server traffic (both in and out)
  3. Monitoring the results of server use (keeping logs, statistics, and analysis)

Within the three areas, the products that monitor servers also cover (albeit somewhat unevenly) a great deal of functionality, which can be broken down like this:

  • Physical: Monitoring the physical hardware includes keeping an eye on the temperature, power supply, and the functioning of components, such as disk drives. Many of these are critical elements, and failure means a dead server. Software that monitors the hardware can be very specific, for example, it works on IBM servers but not Dell servers.
  • Server Performance: Monitoring the performance of a server (e.g., CPU usage, available disk space, and memory availability), especially under a variety of conditions, helps with both troubleshooting and optimization.
  • Services: All servers run a number of services (e.g., DNS, POP3, and TCP). Many of these are critical to server operation. Again, if they fail, the server fails. Most monitoring software covers a wide range of services.
  • Network: An old and very large area of server monitoring is associated with operating a network. This is often considered a separate category of monitoring software, although such functionality is often built into general-purpose server monitoring tools.

In addition, many server monitoring tools are designed for a particular type of server (e.g., Web or database servers). We've provided at the end of this article a ServerWatch Functions Checklist for server monitoring tools. Although the matrix attempts to cover features generally available, it barely scratches the surface of the more-specialized features for monitoring Web servers or networks.

In all, server monitoring software is very diverse, and literally hundreds of products are on the market. Most offer "real-time" monitoring that displays the current condition of servers along with historical monitoring, which is the record of server performance over time. Server monitoring tools are also packaged in different ways: They are always included in the big server management suites, such as IBM Tivoli or Computer Associates Unicenter. There are a large number of general server monitoring products, such as GFI Software Network Server Monitor and BMC Software Server Monitoring and Management.

Specialized products provide features for specific operating systems (Microsoft Windows being an obvious example) and types of servers. To further complicate the choices, server monitoring tools can be purchased and operated by the user, hosted by a third-party company but operated by the user, or fully outsourced (i.e., hosted and operated by a third party). A cursory product search on the Internet will reveal scores of hosted and outsourced approaches.

>> Server Monitoring Tools Matrix

The server monitoring tool selection process is often a matter of matching several things: hardware, operating system, functionality (e.g., Web or database,), and organizational priorities (e.g., preferring software that provides a strong reporting and analytical capability). Some enterprises may opt to match their hardware vendor with the monitoring software (e.g., an xSeries shop may choose Tivoli); others may look for specific features or best-of-breed characteristics in different kinds of servers (again, most commonly for Web and database servers).

And don't forget specific organizational needs. When comparing server monitoring products, it may be important to evaluate the options for alerting administrators, as support for specific methods, such as SMS and e-mail, can be crucial for some organizations. Alternatively, the ability to take remote corrective measures, such as server shutdown or script execution, can be of great practical value.

The ServerWatch Server Monitoring Tools Matrix below is designed to help in the comparison of products by listing a range of features and placing the servers side by side. Keep in mind, however, that to create a usable matrix, we had to limit the details for several monitoring functions. Hence, the blank cells at the bottom of the table, which are there for you to fill in with any requirements unique to your organization.

Server Monitoring Tools Matrix
  Product 1 Product 2 Product 3
Real-Time Monitoring Physical Status Temperature      
Chassis Integrity      
Power Supply      
Fan Speed      
Server Performance Hard Disk Utilization      
   - Files Open/Owner      
   - File Existence Monitor      
  - File Size Monitor (e.g. Log files)      
Memory Utilization      
CPU Utilization      
Processes (Count)      
Services Monitoring DNS      
SMTP, POP3, or IMAP      
Custom Port Monitoring      
Server Types Supported Web      
Network Monitoring Network Throughput      
Current Logons      
Failover Monitoring      
Other Network Monitoring Points      
Web Site Monitoring Hit Rate      
Page Content Verification      
Database Connection Verification      
Cache Rate(s)      
Security Intrusion Monitoring      
Login Error Monitoring      
Other Security Monitoring Points      
Administration Dashboard (or Other Overview Display)        
Remote or Internet Monitoring        
Monitor Views Domain      
User-Defined Group      
Alert Modes E-Mail      
Phone (Land Line or Cell Phone)      
Network Alert      
Escalating Alert Levels Support        
Alert Multiple People        
Designate by Type of Alert        
Monitoring Interval        
User Configurable Monitoring Intervals        
Corrective Action Support Machine Shutdown or Reboot      
Service Shutdown or Restart      
Process Shutdown or Restart      
Run Script, EXE, and Job      
Testing Suite — Automatic Testing Support        
Historical Monitoring Logs        
Reporting Online      
Charts or Graphs      
This article was originally published on Wednesday Jun 15th 2005
Mobile Site | Full Site