About systems management

Computers and servers have become a necessity in all businesses. The number of servers has increased in corporates, schools, and other environments. At the same time, we must recognize that the systems management by manual operation is a huge risk. The systems management of the system always tends to be put aside when daily routines are busy, because people perceive that it does not directly relate to profit. However, abandoned system and manually operated system are very risky. Business executives must understand the risk caused by the system failure.

Risks when the system is not managed

Unable to understand the current status of the system and the server

→ When the usage of the disk and the resource is not understood, the system trouble will be unpredictable. Trouble will not be prevented.

Sudden failure of the system

→ Stops business activities
→ Stops the IT services in the company (e-mail, WEB services, core system, etc.)
→ Stops internal operations
→ Difficult to find the location of the failure and the cause. Just time will elapse

The system is dependent on the engineer

→ No one understands the operation of the system when the engineer resigns.

Scheduled tasks of the system may be forgotten because it is ran manually

→ Backup tasks of the system and the data on a regular basis, and daily batch processing of mission-critical tasks
→ Daily batch application tasks

When these tasks are forgotten, there will be a disturbance or a stoppage in the daily operation, or big problem may occur with the server security.

Human operation error

→ Since the engineer directly access the server, there will be human error. Important data may be accidentally deleted.

Malicious operation to the server

→ When direct access to the server is allowed, there is a risk that malicious operation may be performed on the server. Personnel that can directly access to the server and perform operation must be reduced.

Limitation in manual system systems management

The system operation can be performed manually in the early stage. However, manual operation consist many problems. The thought that "it works, so it is fine" is no longer an excuse.

Manual operation by the script increases cost

→ The systems management can be performed by creating scrip manually. However, this requires creation of many monitoring commands. The development cost and the time to fix that development will add up each day. Also scripts may be recreated each time the personnel in charge changes or the created script may be eliminated due to lack of management.

→ It is difficult to centralize management and reduce labor costs when the system operations management is performed manually.

The system must grow out of the engineer dependency

→ Manual system operations management enhances the engineer dependency. After sudden retirement of the engineer, there is a risk that no one understands the system operation.

→ Hiring skilled engineer will increase the cost of labor.

For end users, the computer system is the use of the service. In other words, the value of the system are the contents and the quality of the provided services. The purpose of the systems management is to provide high level of service.
Based on this perspective about the systems management, the function of the operations can be divided into two parts.
The systems management must be automated as much as possible to reduce mistakes and to create efficient and labor-saving management.

Functions that monitor normal or error operations in the system

Log monitoring

By monitoring the log, it is possible to recognize the cracker activity and the error due to hardware error or incorrect configuration.

Network monitoring

Periodically executes ping to the server to confirm the network connection and the server activity. Also periodically monitors the amount of traffics and distinguishes appropriate circuit capacity by taking statistics of the time zone with concentrated access.

Service monitoring

Monitors connections of the network and the network service on regular basis to confirm normal communication.

Performance / Resource monitoring

To prevent communication failure, periodically monitors the operating process counts and the traffic volume of the server with high loads. If anything did happen to the server, periodical monitoring can provide a hint to solve the problem.

Administrative tasks to be performed periodically at the system

Job management

If there are multiple servers, batch operation of the systems management tool will reduce the workload of the administrator and the operator.

For example, backup and log collection can be performed automatically by the standard features of the OS (cron, task scheduler, etc.). However, when the number of servers increase, it will be difficult to understand which server has what type of scheduled batch processing. This can lead to late discovery of an error and increase the risk of further trouble.

By introducing a systems management tool, it is possible to execute a batch or a job to multiple servers and to execute routine tasks such as backups and night batch. Job management feature provides efficient service to all users.

Batch control

Configuration changes, patch application, and software installation are daily administrative tasks and they need to be performed efficiently. Batch operation by the systems management tool can execute processes to multiples, execute patch applications, and operate startup/shutdown.