Home 
 1. Overview  
 What is preventative maintenance? 
 Why use preventative maintenance? 
 Who uses preventative maintenance? 
 How does preventative maintenance work? 
 What does preventative maintenance cost? 
 2. Preventative maintenance implementation guide 
 What needs to be done? 
 Who does it? 
 Prepare to implement  
 Implementing Preventative maintenance  
 Reviewing the implementation 
 Resources 
 3. Preventative maintenance operations guide  
 What needs to be done?  
 When does it need to be done? 
 Who does it? 
 How is it measured? 
  Resources 
 4. Toolkit  
 Installing a firewall 
 Installing antivirus software 
 Developing fail safes 
 Implementing a backup and restore process 
 Maintaining computers (Apple) 
 Maintaining servers 
 Maintaining switches, hubs and routers 
 Setting up a maintenance schedule 
 Setting up RAID 
 5. Review of Preventative Maintenance 
 Recap of Preventative Maintenance 
 Checklist 
Maintaining servers
Daily server maintenance
Your server is the most critical component on your network, often accomplishing many tasks such as:
  • ensuring that printers are available to everyone on the network, and regulate printing so that it happens in an orderly fashion
  • providing virus protection to all clients on the network
  • sharing information from central CD-ROMs
  • creating nightly back-ups of important information stored on servers or clients
  • monitoring network traffic and warning of impending trouble
  • displaying information from shared databases such as student records, lesson plans, attendance information, or lists of web links
  • ensuring that only authorised users can see sensitive information
  • providing a centre where applications can be stored, updated, and delivered to clients.

It is therefore very important to make sure your servers are in good health at all times.

One aspect to consider is whether your server is working too hard. You may want to consider installing and running more than one server and assigning specific tasks to specific servers if:
  • some of your application programs require a great deal of processing power (that is, they require a server all to themselves in order to run efficiently)
  • your network is very large (perhaps more than 500 clients) and includes many different kinds of network services such as email, access to the internet, web publishing, and databases (accounting, student records, or computer- based instruction systems)
  • your network includes extremely sensitive information that would be protected best if it occupied a server separate from all others (unauthorised users can be denied access to everything on this server, or even to the part of the network where it resides).

Whatever your server configuration, servers that run reliably have one common characteristic: someone watches them like a hawk. Someone knows their baseline performance, detects any variation from that, takes appropriate action to ensure undetected incidents don’t become problems and logs any actions taken.

For the systems you believe to be critical, make sure that someone performs the following preventative maintenance every day. At least two technical staff should know how to perform all preventive maintenance procedures.

Go back to Toolkit
Daily server maintenance
Although each type of server requires different daily procedures, certain tasks are common to all server types.
  • Use the server logs. Servers keep logs of internal errors. Technical staff should be familiar with the location of all these logs, check them daily, and attend to any errors immediately.
  • Use the server statistics. Servers also keep many different kinds of usage statistics such as collisions among data packets on the network, percentage of server  processor utilisation and peaks in network traffic. Technical staff should become familiar with how these statistics look at different times of the day and be suspicious of any unexplained variation from normal behaviour. They should take action to keep the machine at its baseline performance.
  • Check the system daily. Major software systems keep logs of their activity and errors, as well. Technical staff should check each of these major systems daily. These systems include email, web servers, accounting programs and student records databases.
  • Keep a record of error messages. Whenever errors occur technical staff should print any error messages and keep them permanently in a log book. If they take action to address an incident or problem, they should record the date of the action, their names, and a summary of the steps taken. Over time, these log books provide invaluable troubleshooting and support tools. Technical staff can look back to see how they solved an incident or problem, or how one solution caused another set of incidents or problems.
  • Restart the server. Most servers need to be restarted periodically. In part, restarting clears hidden difficulties in memory that may rise at an unexpected moment to bring your server down. Additionally, restarting exercises all the components of your system such as hard disks, memory and network connections. Restarting tests a computer and sometimes forces a failure that otherwise might occur at a more inconvenient moment. Many schools bring their servers down once every month or two.
  • Test all procedures. Support staff should simulate procedures for starting servers in an emergency. Most servers provide diagnostic tools for booting (starting) and helping to identify the source of incidents and problems. The tools for booting include special disks or tapes. These disks and tapes must be prepared ahead of time, when the server is healthy. Technical staff should know the location of these boot disks/tapes, and they should test the boot procedures periodically. Most disks and tapes degrade over time, and many sites recreate their boot disks or tapes periodically to refresh them. In addition to emergency boot procedures, servers provide diagnostic software. Staff should know how to locate and use this software.
General guidelines for server maintenance
You may find the following general guidelines helpful.
  • Keep a list of all system passwords in a secure location. All support staff should know the location of passwords. They should keep the list up to date. Make sure that passwords to privileged accounts (system manager, administrator and so forth) change frequently.
  • Post and maintain a list of emergency support telephone numbers, customer identification numbers and all additional numbers required to secure support.
  • Have a support contract from the manufacturer for your most complex systems (accounting, student records and so forth). While your ICT staff should still have thorough knowledge of all your equipment and be responsible for its daily care, outside experts can be invaluable in times of trouble.
Top