Network issues and disruptions can cause a loss of productivity. Customers have to deal with undesirable lag and poor network performance. Although some of these issues may be easy to spot and resolve (or are outside of the data center and it’s bandwidth carriers) there are times when these networking issues may take more time and longer to troubleshoot.
It is important to respond and investigate client feedback or complaints regarding networking lag and slowdowns in order to determine whether it is an ongoing and escalating issue or a specific issue related to just one client. If monitoring is not in place or not detecting slowdowns, client feedback is a vital part of determining less than optimal network speeds. If user complaints are vague, hosting providers should have technical support step in with additional questions regarding network issues. Without details, it will become difficult to determine where the issue is originating.
Troubleshooting network issues involve:
what time of day the issue occurs
traceroute results
what apps or software they are running
expectations from clients
bandwidth carrier status reports
Obtaining feedback
Obtaining feedback from several clients can also provide vital information on patterns with the network. Especially if clients are reporting similar issues, it can be easier to track down and resolve. Several different complaints will cause more investigating and followup questions. If complaints are all coming from the same server running virtualization it can be surmised that the issue is with that particular dedicated server connection. Feedback coming in regarding various networking issues outside the data center, usually means the broadband carriers are experiencing issues. A notification system to clients via some sort of online bulletin board can help lessen the concern in this case.
Reactive technical support helps speed up the problem solving as well as wait time for networking issues to resolve. Planning ahead and foreseeing possible network congestion during company growth is a must-do in keeping clients happy.
It’s not the network
In many cases what may seem like network issues at first can very well be the hardware and dedicated server. There needs to be enough “horsepower” in a server to keep software running, especially in a virtual hosting environment where oversubscription is a common occurrence in the web hosting industry. Fast networking and low latency datacenter have no effect on an overburdened dedicated server that cannot keep up with requests. Doing the right math on this is vital. Older servers may simply not be up to the task of running newer applications and software available today. These newer apps expect more CPU power and RAM.
Update!
Firmware updates for lots of networking hardware are essential and can be part of the blame for network issues. Vendors release such updates on a regular basis. This can be security related or performance upgrades, resulting in network improvements. This can be especially true for NIC’s and routers/switches. Visiting a vendor’s website for patches, updates or the latest drivers can help any data center network. Especially if a new Operating System upgrade has been announced. Hardware vendors need to be kept up-to-date every time a new OS is released to ensure no issue could arise at the time of the OS upgrade.
There is also the issue of course that the data center network is simply not meeting client demand due to hardware no longer being sufficient to keep up with growth. This can result in a major network upgrade. Knowing what current trends are and what future plans should be will result in smaller growing pains for any datacenter or web host provider. As more services such as VoIP and video/audio streaming continues to grow, so will bandwidth requirements.
Monitoring network issues
A Network Management System allows for traffic monitoring such as when peak traffic consumption occurs and when applications are grappling for the same bandwidth as well as lag times and choking points. Determining whether a client is consuming too much bandwidth is up to the web hosting provider. They will need to decide whether the client(s) have broken the Acceptable Usage Policy. Upgrading a network based on a small number of clients may simply not make much financial sense.
Not all clients, however, will show displeasure in less than adequate network speeds and will simply cancel services leaving IT techs and management scratching their heads as to why. Cancelation forms with several options as to indicate the reason for cancellation is especially helpful here.