The use of MySQL in the cloud is broadly divided into two categories:
IaaS: Infrastructure as a service, IAAS provides the basic services for your MySQL server, you can purchase virtual server resources to install the MySQL server instance. You can set the OS and MySQL server according to your own requirements, but you cannot relate to the hardware facilities.
DbaaS (Database as a Server)
MySQL itself is the cloud management resources, you have only one access to MySQL credentials, you can only set up MySQL, but do not see the OS and virtual resources. Amazon's RDS is an example of a service that is not really MySQL, but compatible with MySQL and query language.
Most of us focus on the first category. (Our goal is to help you avoid some of the pitfalls you may encounter if you are not a MySQL expert in the cloud.) )
Pros, flaws, misconceptions about the cloud
Cloud computing has many benefits, some of which are specific when using MySQL.
Advantages:
1, all the infrastructure is outsourced, without management and trouble, you do not have to buy hardware and develop supply chain relationships. No need to replace broken hardware, etc.
2, pay is according to their own needs, can be the early capital investment into the current project operations.
3. The value of cloud servers increases over time as they provide new services and reduce costs. You don't have to do something to upgrade your services to benefit from their value.
4, the Hardware resources management is convenient, may add or withdraw any resources at any time, eliminates the processing or the recovery cost.
5. The cloud brings a rethinking and positioning of the infrastructure, and resources can be presented through APIs and more automated. You can also build your own private cloud.
Disadvantages:
1, the resources are shared and not fixed. The resources we get are not as fair to use as we pay for, and the cloud server doesn't give us a satisfactory answer.
2, for the business capacity and reliability is not guaranteed. Can not completely according to their own business needs to purchase new resources, may receive supplier resource constraints or "excess mergers and acquisitions"
3, virtual, shared resources is not good for troubleshooting. In particular, you will not be able to see the relevant hardware facilities and know their status. For example: When we look at the system, Iostat display IO Normal, vmstat display CPU normal, when we really test found the problem is because other programs on the system caused the system load high. It's really bad to judge when a problem arises.