What levels are commonly used in raid? What are their respective features? What are precautions for using raid in Oracle databases?

Common levels of RAID are 0, 1, 3, and 5. The combination is usually 0 + + 0. Their respective features are described as follows:
Raid0: Multiple physical disks are made up of a large Logical Disk with the total capacity of each disk. RW has the best performance, but the problem is that if a disk is lost, the entire raid data will be lost.
Raid1: an image. The data in multiple disks of RAID 1 is the same. In this way, the capacity of multiple disks is one disk, which provides data protection and good read/write performance, in some cases, the read performance is about 15-20% higher than that of a single physical hard disk.
Raid3: one of the multiple raid disks is used to store verification data. When one non-validation disk is damaged, the data is automatically restored to the new disk. However, when the verification disk is damaged, the entire raid becomes invalid, that is, there is a single point of failure. The write performance is poor, but the read performance is good, because when reading data for an IO request, all the disk heads are located in the same position (the same cylindrical disk, the same sector)
RAID5: similar to raid3, verification data is distributed across all hard disks on the entire raid. No matter which disk is lost, data on the raid is not destroyed. This method has poor write performance and good read performance. It is also suitable for scenarios with frequent random reads and writes, that is, multiple hard disk heads can independently serve different IO requests, improving Read and Write Performance.
1 + 0 is better than 0 + 1, the reason is that when 0 + 1, if one of the 0 disks is damaged, these half of the logical disks are invalid, in theory, the read performance is 50% lost (refer to raid1). When 1 + 0, if a disk is lost in 0, its image still exists and can work normally, the entire 0 still exists, with little impact on performance.
When using raid in Oracle, consider the following:
Two Disks are used as raid1 to store online logs. It is best to open three volumes and store three groups of two members in each group.
Two Disks are used as RAID 1 to store archived logs.
If you cannot create a dedicated raid1 for archiving, you should also open a separate volume group for it to put it on the same raid1 as online logs.
The other disks are used for raid3 or RAID5. Which of the following depends on the purpose of the application? For details, refer to: raid3 is better if the DB application is data mart/Data Wh; RAID5 is better for general OLTP applications. For details about this principle, refer to the features of raid3 and RAID5.
In addition, in the partition of tablespaces, the temp, RBS, and system tablespaces are best distributed in the same volume group, with the least competition, in addition, data table spaces and index tablespaces must be stored in different volumes. They are usually the main reason for competition.