There are many elements that affect network traffic, among them bandwidth, application and utility scheduling and hardware such as routers and network cable. Network traffic can even be slowed by employee errors such as all-employee broadcast messages sent during peak production hours. But in any analysis of a “slow network”, it is worthwhile to check fragmentation–and the defragmentation technology in use.
How does file fragmentation make a network appear slower? Let’s take an example: User A accesses a file from Disk B on Server C. If that file is fragmented, it will take longer for Server C to collect all the file fragments from Disk B and transmit the file back across the network to User A. User A, sitting and waiting for the file to appear, may very well think that the network is slow for some reason. The user might even complain about it, causing the IT department to begin an investigation into the “slow network” when the problem may not lie with the network itself at all.
This scenario could extend to email; if the email server has a fragmented drive, a user will have to wait longer for email while it is retrieved from the hard drive, creating the same type of illusion–that the network is slow. Fragmented drives on a database server could also lead to the same incorrect conclusion.
Most enterprises employ scheduled defragmentation as a solution to the fragmentation problem. But many companies, especially in the last couple years as file and drive sizes have grown so dramatically, have complained that overall performance still slows down.
This is because scheduled defragmentation, while once adequate to the job, can no longer keep up. In between scheduled runs, fragmentation is continuing to build, impact performance, and create the “slow network” impression. And in some cases, as with very large drives, defragmentation may not be occurring at all. Many system administrators and IT personnel may well be “chasing ghosts” in an effort to solve slow networks while hidden file fragmentation is still causing the problem.
Companies are now catching on that the best solution for defragmentation today is one which is fully automatic and runs invisibly in the background whenever idle resources are available. Dives are kept constantly defragmented and performance is always maximized.
With a fully automatic defragmentation solution in place, you know that if the network does slow down, it really is the network.
