A system administrator’s job is tough. Dealing with finnicky hardware, obtuse software, and wailing clients can get you down. You need all the help you can get! Here are some tips to make your day go more smoothly.
It goes without saying that you need some kind of trouble-ticket system. You need software to track the flood of requests you receive. Systems administrators (SAs) receive too many requests to remember them all. A system documents what requests are received when and by whom. It also documents client agreement that the request has been completed.
If you’re the only Sys Admin is a small company, your trouble-ticket system could be a simple Microsoft Excel spreadsheet or Word document using a template. When you receive a request, talk to the user to get a sense of the scope. After logging the request in your “system”, send an email or text (depending on complexity) outlining the request. Include an estimate of completion and ask the user (or the user’s manager) to approve the request. Believe it or not, this formal procedure will cut the number of requests significantly. Users often make off-hand requests without considering the SA’s time and effort. When they are confronted with the request in black-and-white, they often withdraw it.
In a larger company, a trouble-ticket system could be a bells-and-whistles helpdesk software program. The system tracks not only the request, but also to which Sys Admin or IT support person it is assigned. Remember that the system should include all requests from IT personnel as well, not just from users.
There is an important by-product of systematizing requests; the Sys Admin now has documentation of his/her activity. If a manager asks, “What have you been doing?”, just pull up the trouble-ticket system and show your accomplishments. Even better, send your manager an email with the number of requests received and completed each week.
Many network and device problems can be traced to power fluctuations and overheating. An Sys Admin’s time can be eaten up by chasing these transient problems. So make sure that each room or closet containing routers, switches, or servers has sufficient power and cooling. Install uninterruptible power supplies (UPS) that can survive a 1-hour outage and gracefully shut themselves down before the batteries die. Since most outages last only a few seconds, this measure will drastically decrease the number of time-consuming device reboots required.
Cooling is even more important than power. Smaller organizations often house their network devices in closets with no cooling. This is an invitation to disaster. Minimally, install a portable air conditioner that vents through the ceiling. This has an advantage over a central air hookup: many building maintenance systems stop or limit air conditioning after hours and on weekends and holidays. In this situation, a server that works fine during the week can overheat and shut down on the weekend.
Do your users think that all servers tend to crash on Monday morning? Then you have a monitoring problem. Crashed machines accumulate all weekend, and the Sys Admin is left with a mess on Monday morning.
There is no excuse to skip monitoring: there are too many simple, inexpensive solutions. For any business with a monetized web presence, comprehensive monitoring is a business requirement because a system outage slashes revenue. If you have many remote locations, monitoring connections is essential, especially if there is a service-level agreement (SLA) with your Internet provider. Depending upon network configuration and complexity, this could be as simple as using the SLA monitoring function in your router or switch.
Know how much of your total bandwidth is consumed by monitoring applications. The consensus is that one percent of available bandwidth is the limit. For smaller businesses, if you have a few servers and no remote locations, simple text alerts should suffice.
It is important to “monitor your monitoring system”. It is great not to receive alerts, but how do you know that your monitoring system hasn’t died?
It is amazing how much time SAs spend repeating the same tasks over and over. Whenever a task is accomplished manually, there is plenty of room for error. There is a better way. Use the automated features inherent in servers and other devices for installation and updates. Most devices offer scripting capabilities as well. For example, Windows 2012 Server support is streamlined if you use Powershell.
Why is this important? Consistent configuration means fewer helpdesk calls. With automated updates, security is improved. There is a learning curve for scripting, but it is worth it. You will bring devices online faster in the future, and users will be happier.
For a very small business, at least have a checklist with detailed instructions including exactly what options and preferences are to be set on various applications.
I know what you are thinking. How? For 24/7 operations and/or larger networks, there is undoubtedly redundancy. The backup devices can be upgraded, their fiber connections cleaned, and so forth. There is always a new software release to be tested. If you schedule regular maintenance windows with rolling outages, there will be better availability during normal operations.
It is difficult to communicate technical subjects to a non-technical audience – the users you support will often be non technical. It requires practice, patience, and, above all, listening to others. How well you communicate with a user can determine the perceived success or failure of a project.
Use a login banner to advise all users about maintenance windows, for example. Create a website with the helpdesk number, an email address, and a list of who to contact for various tasks. Include a description of the tasks; users do not know the difference between “Server SA” and “Network SA”. The website should be the go-to place for users to find out if there are any changes such as upgrades affecting specific systems or applications. It is a good idea to have some self-help options like how to change your password or restart your computer. This can lessen the load on the helpdesk.
Finally, it is good public relations to let users know when a system problem has been fixed. Announce this on the website and/or via email.
Bottom line, every employee must make or save money for the business. An SA tends to concentrate on technical issues at the expense of the big picture. But it is critical for your success to understand how the business works. Especially in smaller firms, the manager authorizing IT expenditures may or may not be up-to-speed in IT matters. It is your responsibility to explain the business reasons underlying requests for software, hardware, or modifications.
It’s up to you! The seven habits above cover the three main areas that a system administrator must get to grips with to be good at his / her job. These are technical, interpersonal and procedural areas. All of these areas are equally essential to your success in the role.
Expectations are always high when it comes to the performance of the system/ network administrator. If you develop and hone the skills we have discussed this can help turn an average system administrator into an extraordinary network hero.
Sign-up for email updates...