Below is a list of some of the more common tasks that a system administrator may be expected to perform. In real life, for any given position, a system administrator may be required to do more or fewer tasks than those listed here.
Your personal list will nearly always be unique to your level, experience, situation, goals, and values. A typical day might look like this:
|Arrive at Work||Check service desk for new and on-going requests and issues|
|Morning||Check monitoring system(s) for new incidents and the status of recent incidents and changes|
|Work on service desk request(s) and issue(s) |
Work on emergency that “suddenly” came up
|Early Afternoon||Attend meetings (future planning, management reporting),
make vendor and other calls
|Late Afternoon||Work on service desk requests
Self-training and advancement preparation
|Update status and work on reports|
The service desk is the main application where users, help desk personnel, and system admins can create requests (for new features) or report issues. This software is often known as a “trouble-ticketing system”. Example requests include new employee (create login, email, and other IDs, setup access, etc.), new project (create new website, new database etc.), upgrade services and software to new versions, upgrade hardware (memory, disks, etc.), and so on. Examples of issues include login/access not working (website, FTP, etc.), site (website, etc.) not working, printing issues, migration of old website or other data to new service, backup and restore requests (this is vital, and you can't afford to mess it up), employee termination, and so on.
Monitoring is usually a simple matter of examining reports from monitoring software, or (cheaper) examining log files. Over time, you must adjust what is monitored. Often monitoring is automated to some extent by crafting shell (or other) scripts to do most of the work for you, and to generate simple reports and alerts. Some simple checks which can be done daily include uptime checks for each host, log summaries (or sometimes raw logs: dmesg , maillog , secure, messages, etc.), network status (including bandwidth used and error counts), hard disk (storage volume) free space and errors, backup status, and various performance metrics (such as CPU utilization). Monitoring is related to capacity planning; you need to be able to predict when to order upgrades, replacements, and new equipment and services.
Another routine task of a system administrator is planning for the future. From time to time, think in a quiet place, talk with colleagues, talk with users/customers, and understand where you may improve your sites, services, and procedures. Then, research technologies and “best practices” that may be used (especially new ones) to decide which may be worth using. Research may involve proactive project creation, some project management, predictive analysis, and statistical analysis. Then, if necessary, get management permission (and budget) for implementation of these research (or “pilot”) projects. You must also find time to keep up your skills, as well as building your career.