sparse disks

[Roshan]

Hello Erman,

kindly advise how the virtual disks become 100T while the size of physical grid disk is 4T on 1 cell?

https://blogs.oracle.com/exadata/post/exadata-pdb-sparse-clones

Regards,
Roshan

[ErmanArslansOracleBlog]

Just some background info for sparse clones.. This may be due to tech used there.. It is like copy on write snaphots..
Sparse Clone (copy-on-write).. That is, The master of a sparse clone is read-only. **Sparse clones contain changed blocks Blocks accumulate as the clone changes...***"

For the sparse disks, we have 2 concepts.. Physical size and Logical size..

--Calculating the Physical Size for Sparse Grid Disks:


Total physical space =
(SUM(size of all test in the sparse ASM disk group) +
SUM(approximate size of all updates to the snapshot databases))
* ASM Redundancy
 

Example:

Space for one test: 15TB = 15 TB
Space for 2 snapshots per test, for a total of 2 snapshots:
2 * 15TB * 20% = 6 TB
Subtotal 21 TB
High redundancy: 3 * 21 TB GB = 63 TB

--Calculating the Virtual Size for Sparse Grid Disks:

Virtual size required for sparse disks =
(SUM(full virtual size of all Exadata snapshots) + Physical space allocated)
* ASM Redundancy
 
Example:

Full size for 2 snapshots per test, for a total of 2 snapshots:
2 * 15TB = 30 TB
Size of one test: 15 TB = 15 TB
Subtotal 45 TB
High redundancy: 3 * 45 TB = 135 TB

Ref: How to Calculate the Physical Size and Virtual Size for Sparse GridDisks in Exadata Sparse Diskgroups (ORA-15041) (Doc ID 2473412.1)