ASM disk High Redundancy with Ten Disks !

My last question related with ASM Disk high redundancy . I have some confusion on this area .

Let's say I have 10 disks on a diskgroup '+DATA' and I have redundancy level 'HIGH', means 3 way redundancy . Here I know 3-way redundancy means , it writes data on 3 disks at a time . But what about other 7 disks . Actually I need to know the writing mechanism ?

- Does each disk is identical with other disk ?

- For HIGH redundancy, How many disks failure can be tolerated for any data loss  ? 

ASM Diskgroups with NORMAL or HIGH redudancy are further divided into failure groups. These failures groups are logical groups of disks that each mirrors asm file extents on other groups. The mirroring is done at asm file extent level, not disk level. So a disk will containt asm file extents which are primary copies and also copies that are secondary mirrors of other disks. A nice visual explanation of this concept is found here: Brief introduction to ASM mirroring | The Oracle Instructor

site : http://uhesse.com/2015/01/15/brief-introduction-to-asm-mirroring/

Oracle doesn't mirror entire disks.  As Alvaro has explained, Oracle mirrors extents.  With High Redundancy, an extent in 1 disk is mirrored to two different failure groups.  If you have defined your disks as 7 distinct failure groups, an extent from disk 1 may be mirrored in any two of the other six disks -- say disks 3 and 6.  The next extent in disk 1 is, again, mirrored to any two of the other six disks, possibly disks 4 and 7. And so on .  So no one disk is entirely mirrored to another disk.  It is a small portion (the extent) that is mirrored. Two extents from the same disk may be mirrored to two different other disks.