Author: Westech | Date Published: 06 July 2009 | Article Release

It is difficult to look at a network briefly and diagnose what is causing the problem as it often involves getting to know a network over a long period of time. However this is not always an option here but these are the areas of the network that could be causing issues.

1. Network Cabling

• If electrical cable is near the Cat 5 cabling. Electrical interference could interfere with the quality of the data being transmitted.
• Wiremaps : If the four pairs of the Cat 5 cabling are not done according to specification or in the pairs are not in the correct order as often the case with inexperienced cable companies or individuals: Line attenuation could exist, causing data corruption.
• If cables are severed or bent/squeezed too sharply around conduit corners, data corruption could be caused.
• Insufficient Earthing or cabling near Telecom cables could cause bridging or lightening surges. Causing network cards or active equipment (Switches) to become faulty or unstable.
• It is not advised for wireless to be used for high process intensive applications as:
  • Wireless can be influenced by other wireless nodes in the area
  • Topology of land can affect it
  • Structure of building can effect it
  • iSecurity can be more easily breached

2. Network Topology and Active Equipment

• Switches/ hubs can have faulty ports
• Switches could be cascaded or double linked, or other faulty devices causing network collisions or poor network performance.
• Routers could be allowing unauthorized entry and unnecessary excess traffic causing poor network performance.
• Network Protocols could be incorrectly configured.
  • TCP/IP could be set to an incorrect subnet.
  • A broadcasting protocol like NetBEUI could be installed and used causing network to be slow.

3. Power Quality

• If the server or workstations do not get shut down correctly when there is a blackout. Data corruption could result.
• These blackouts could also cause hardware to become faulty such as server or workstation motherboards and hard drives. This defectiveness could result in data corruption.
• If cheaper offline unknown brand UPS units are used on servers and workstations, these units can create bigger problems, as when power dips they kick in and cause power fluctuations.
• Regular maintenance of UPS batteries. These need to occur every two to three years, old batteries cause problems that they were meant to eliminate.

4. Server Hardware

• If numerous power fluctuations have occurred or power has failed on a regular basis. Then motherboard circuitry could become unstable and in time the server could fail.
• Hard drive surfaces (where all the data is stored) could be faulty caused from knocks power fluctuations or failed hardware, causing bad sectors to occur and data to become corrupt.
• Has Network growth been predicted and made allowances?

5. Operating System Health

• Every time a server does not shut down correctly then Operating System failures occur which in turn creates unreliability
• The more software installed on the server the more likelihood of the server to fail and the OS to be come unreliable and unstable.
• Every time the server fails data corruption can occur.
• Hard Drive Space or the shortage of it, could cause the Server to become unstable.
• Windows Updates are not being done on a regular basis to avoid security exploits and OS flaws.
• Has a Service Level Agreement been put in place? Is there a level of supervision of the onsite technicians? Who do they report to?

6. Antivirus Software or memory resident programs

• Certain memory resident anti-virus programs can cause the server’s or workstations’ performance to deteriote.
• Constant anti-virus scheduled scans during operating hours can also cause issues to occur. These schedules should rather happen after hours.

7. Workstation Health

All the below items could influence the way data is written to the database and effect data integrity.

• Quality of Power
  • Grade of UPS (online versus offline)
• Reliability of hardware
• Operating system
• Anti-virus
• Quality or state of network interface card or cabling.
• Windows Updates are updates being done on a regular basis to avoid security exploits and OS flaws.
• Application updates must be done regularly to update application software and avoid exploits from within its code.
• Has Network Growth bee considered?

8. Environment of Server

• Heat
  • Temperature could cause the server to hang or not respond. The server should be in an environment where temperature is kept constant. This ideally should be 18 degrees for South African conditions.
  • Server Case should be on, sealed and secured. Internal fans need to be in operation so that air is pulled from the front of the server to the back.
• Cabinet should be away from being bumped or obstructed
  • All equipment should be in a self contained rack where it can not be knocked or moved.
  • Cabinet should be constantly tidy.
• Dust Free from falling dust (A cabinet could provide this)
  • All equipment should be in a self contained rack where it can not be knocked or moved.
  • Air movement should be encouraged upwards by top mounted fans. This ensures that does not settle on component circuitry.
• Earthing
  • Cabinet should be properly earthed as to allow surging to dissipate and proper lightening protection to exist.
• Flooring or surface that servers are sitting on must be free from Electrostatic Discharge (ESD).
  • This goes for any surface where computers are worked on in this environment.

9. Internet Access and connectivity

• Internet Access can affect workstations if a virus comes through from the internet and infects the workstation.
  • This can effect network performance to be affected
• Internet Browsing should not be allowed from the Server’s Browser as Spyware can copy itself onto the server and cause performance/bandwidth issues as well as reveal information about the server/network to hackers outside of the network.
• High volume of internet usage can affect network speed.
• If server is infected with Spyware or Anti-virus server can become unreliable.
• Email scanning and Spam removal. Human engineering is the most successful way of getting ino a network or informing outside hackers about your information. Most of these exploits come into the network as an attachment to a Spam email. It is important to have an anti-spam and attachment removal policy and system to implement it.

10. Risk of theft or Disaster Recovery plan. Backup and Corporate Compliance.

• Server needs to be located in a place that is inaccessible to potential thieves. And access must be restricted or monitored.
• Rack mounting is always a good idea as it makes it difficult to remove and discourages it from being stolen as it causes delays. If it appears as a server it also is less likely to be stolen.
• Workstations are always at risk of being stolen as they are in accessible places. Thin clients are always a good option in areas of high accessibility as the likelihood of theft is higher.
• The time needed to be up and running in an event of a disaster should be considered. Backups, Operating System volume and shadow backups need to be considered.
• A Service Level Agreement should also be considered whereby your Service Provider can make sure he has resources on call to respond to your call within the time needed to be up and running. It is highly advisable that this service provider is responsible for your backups.
• Have Archiving Policies Been Provided? Laws are being passed to ensure that companies keep data for longer periods of time. Does he company comply on corporate governance should an audit be issued?
• Data Growth and reliability. Is the site growing? And if so has data growth been plotted and forecast. Has a budget been set aside for the hardware required and most importantly the managing of this growth?
• Have test restores of the backups been done? Have emergency procedures been tested?
• Service Level Agreements should be in place with hardware, software and network providers.

E-mail sales@westech.co.za for an appointment or information on an audit.