ServersCheck Logo

Monday, February 9, 2009

Server Monitoring Systems

Of all the things you might be concerned about when running a server, cluster, or even a full data center; the chances are good that a server monitoring solution can allay the majority of your fears, cut costs, and reduce down-time.

Server monitoring solutions come in many shapes and sizes and frequently boast a blistering array of hardware and software applications, combined with sensors and communication protocols. They can be needlessly complex, deceptively simple, and everything between.
Hyperbole aside, the tasks of any server monitoring system is three fold:
  • Integrate into your server hardware and software without the need to disrupt that hardware or re-deploy software.
  • Use software and hardware sensors to monitor environmental and networking conditions inside and outside of your servers.
  • Continually report back on server status, providing alerts for critical software, hardware, or environmental situations.
In reality this ideally results in complete environmental and systemic monitoring of your critical services so that you, your potential customers or employers, and even your insurers, are never taken off guard by unforeseen hardware damage or software failure.

Though many network administrators are all too aware of the need for continual network device monitoring, most assume individual, periodic hardware failure to be an inevitable but acceptable rarity. This is correct, and server monitoring equipment can certainly help, but there are far more serious threats to your server hardware, dangers which can take down an entire data center for many weeks and potentially demand the replacement of all server hardware.

Environmental conditions change and, even in the most highly secured custom built server room, there are still a range of risks which can wreak havoc upon your hardware. These come in the form of excess humidity and temperature, power failure or, in the most extreme cases: Fire and Flood.

Server monitoring hardware sensors can be deployed throughout your hardware and environment to keep tabs on these otherwise silent server killers. Admittedly, most data centers or corporate IT offices are equipped with air conditioning that controls overall temperature and humidity, but a simple failure of this single part –while unpleasant for human employees- can spell the end for server hardware.

Enough doom and gloom, now. Any experienced server administrator or technician will understand the damage which can be caused by overheated systems or excess humidity, and hopefully they have implemented a monitoring system to keep a watchful eye on these environmental factors. On the other hand, more pressing and frequent problems are caused by isolated software and hardware failures which, unless monitored closed, can go unnoticed for many days.

The range of software monitoring options can be dazzling and varied, and while most monitoring systems will allow for a range of monitoring options, only a few do so with ease and consistency. Systems being monitored can include a raft of server applications, based upon their assigned ports, these might include:

  • Internet Related Checks
  • SNMP Related Checks
  • Mail Related Checks
  • Media Server Checks
  • OS Specific Checks
  • Applications Specific Checks
  • VOIP Checks
  • Database Checks
  • File Checks

Suffice to say, the built in functionality of server monitoring systems should cover the majority of your needs but, due to the way the software monitoring system works, it should also be possible to set up custom checks on any running service you desire.

These networked software and hardware tests are carried out in three steps. First, the monitoring software needs to know which services you wish it to monitor, this can be configured manually (via I.P. and port specification) or, in the case of ServersCheck, you may be able to automatically map your network devices and services.

In either case, the server monitor will periodically check known IPs and Ports to monitor their response state and times. In the event that any part of the system fails and a response is not forthcoming, your server monitor can be configured to alert you in a variety of ways, such as:

  • SMS
  • MSN messenger
  • Email
  • Voice messages
  • Or audio and video alarms

This is also true for environmental warnings which might need human attention. In this way, there are multiple opportunities to track your server performance and environmental factors, any of which could give you a clue to why a server has failed or, preferably, that a server might fail due to one of the tracked statuses.

The final part in the puzzle is the tracking and display of these environmental and network environments over time. This is where a server monitoring system becomes far more than it might immediately seem. Using ServersCheck system as an example, you can expect graphs which display the following variables readings over time:
  • Server, environmental and network status
  • Temperature and humidity over time
  • Bandwidth usage over time
  • Many more.
A combination of these unique and robust server monitoring systems ultimately ensures you have total control of your server and its environment. When combined with time based reviews, you can gain unique insight and understanding of the many varied factors which contribute to the cost, up-time and failure points in your server. Ultimately, you’ll save time, money and steer clear of potential catastrophe.