A few years ago, Zhao squad layer summed up the operation and maintenance of knowledge: https://www.unixhot.com/page/ops ; for the operation and maintenance system to a comprehensive exposition.
Today, Sohu architect listened to a course, deep feelings. Now combine personal understanding, do the following aspects to illustrate, for any errors, open to criticism.
Personally think that just entered the operation and maintenance work of newcomers, transport Dimensional Thinking breakthrough is important:
(1) Backup, disaster recovery;
Local disaster recovery, remote disaster recovery;
Data backup disaster recovery is an eternal topic, such as guaranteeing useful data traffic generated by the effective protection when the unpredictable catastrophe, to ensure that data is not lost.
Consider admitted from the most basic servers
If you are on the same server, backup data must first be copied, how the data quickly and efficiently copy or replicate in real time on the level
Of course, data backup on the same server, there is a greater risk, when the server fails, all data will be lost, so we need to back up data on different servers, so that there is a problem of data transmission, how to ensure efficient data transfer.
Further, the same data in a storage cabinet or room, or even the same area, the damage may be unpredictable natural disasters or man-made, and thus is generated, the off-site disaster recovery data backup, and data backup for offsite disaster recovery, data transfer link will be longer, how to ensure the secure transmission of process data, complete and implement a backup.
System backup, database backup;
Considered from the level of a single server, system resources backup lecturer grace importantly, Cai Deming predecessors in the "Bird Brother private kitchens" set forth clearly.
The database, we have from the Lord, The Lord and other high-availability architecture, and secondly, for the non-relational databases, we consider how persistent data.
(2) Data Migration
In fact, with the development of technology, data storage is not the traditional local stores unless the operation of the system must be resources. Most of our storage resources based on network storage. Second, with each passing day, the arrival of big data era of the Internet, storing large amounts of data, how fast write and read?
Data migration issues to consider:
Security: How to ensure that the data in the transmission process, not stolen?
Data integrity; how to ensure that data flow is not dropped frames?
Transmission speed; how fast transmission of data, to ensure the timeliness of the data storage while improving the customer experience degrees.
In data transmission, How do we ensure the transmission speed and link redundancy.
(3) management and operational separation
(4) Change Management
In our Internet companies, how to ensure the accuracy of the new service line as well as changes to the service, expeditious and orderly operation.
(5) access control
For high-volume machines, how do we control its authority, and how records related to the operation to be performed on the machine.
(6) Automation
For resources, how do we monitor;
For the link, we how to monitor: