When you configure the Erasure Coding profile for an ILM rule, you select an available erasure coding scheme. Erasure coding schemes control how many data fragments and how many parity fragments are created for each object. The erasure coding schemes that are available depend on how many Storage Nodes and sites make up the storage pool you plan to use.
The StorageGRID Webscale system uses the Reed-Solomon erasure-coding algorithm. The algorithm slices an object into k data fragments and computes m parity fragments. The k + m = n fragments are spread across n Storage Nodes to provide data protection. An object can sustain up to m lost or corrupt fragments. k fragments are needed to retrieve or repair an object.
You can use erasure coding with a single-site deployment or with a multiple-site deployment. Erasure coding is well suited for single-site deployments that require efficient data protection with only a single erasure-coded copy rather than multiple replicated copies.
Erasure coding scheme (k + m) |
Number of deployed sites | Recommended number of Storage Nodes at each site | Total recommended number of Storage Nodes | Site loss protection? |
---|---|---|---|---|
2+1 | 3 | 3 | 9 | Yes |
4+2 | 3 | 3 | 9 | Yes |
6+3 | 3 | 4 | 12 | Yes |
9+3 | 4 | 4 | 16 | Yes |
6+2 | 4 | 3 | 12 | Yes |
8+2 | 5 | 3* | 15 | Yes |
* At minimum, each site requires three Storage Nodes. Additional erasure coding schemes are available. Contact your account manager. |
The following example shows the 6+3 erasure coding scheme, which splits each object into six data fragments and adds three parity fragments. This erasure coding scheme requires a minimum of nine Storage Nodes, with three Storage Nodes at each of three different sites. An object can be retrieved as long as any six (k) of the nine fragments (data or parity) remain available.
If more than three (m) Storage Nodes are lost, the object is not retrievable.