Skip to main content

Example ILM policy for FabricPool data

Contributors netapp-lhalbert

Use this simple example policy as a starting point for your own ILM rules and policy.

This example assumes you are designing the ILM rules and an ILM policy for a StorageGRID system that has four Storage Nodes at a single data center in Denver, Colorado. The FabricPool data in this example uses a bucket named fabricpool-bucket.

Caution

The following ILM rules and policy are only examples. There are many ways to configure ILM rules. Before activating a new policy, simulate the proposed policy to confirm it will work as intended to protect content from loss.

To learn more: Manage objects with ILM

Steps
  1. Create a storage pool named DEN. Select the Denver site.

  2. Create an Erasure Coding profile named 2 plus 1. Select the 2+1 erasure-coding scheme and the DEN storage pool.

  3. Create an ILM rule that applies only to the data in fabricpool-bucket. This example rule creates erasure-coded copies.

    Rule definition Example value

    Rule Name

    2 plus 1 erasure coding for FabricPool data

    Bucket Name

    fabricpool-bucket

    You could also filter on the FabricPool tenant account.

    Advanced Filtering

    Object Size (MB) greater than 0.2 MB.

    Note: FabricPool only writes 4 MB objects, but you must add an Object Size filter because this rule uses erasure coding.

    Reference Time

    Ingest Time

    Placement

    From day 0 store forever

    Type

    Erasure coded

    Location

    DEN (2 plus 1)

    Ingest Behavior

    Balanced

  4. Create an ILM rule that will create two replicated copies of any objects not matched by the first rule. Do not select a basic filter (tenant account or bucket name) or any advanced filters.

    Rule definition Example value

    Rule Name

    Two replicated copies

    Bucket Name

    none

    Advanced Filtering

    none

    Reference Time

    Ingest Time

    Placement

    From day 0 store forever

    Type

    Replicated

    Location

    DEN

    Copies

    2

    Ingest Behavior

    Balanced

  5. Create a proposed ILM policy and select the two rules. Because the replication rule does not use any filters, it can be the default (last) rule for the policy.

  6. Ingest test objects into the grid.

  7. Simulate the policy with the test objects to verify the behavior.

  8. Activate the policy.

When this policy is activated, StorageGRID places object data as follows:

  • The data tiered from FabricPool in fabricpool-bucket will be erasure coded using the 2+1 erasure-coding scheme. Two data fragments and one parity fragment will be placed on three different Storage Nodes.

  • All objects in all other buckets will be replicated. Two copies will be created and placed on two different Storage Nodes.

  • The erasure-coded and replicated copies will be maintained in StorageGRID until they are deleted by the S3 client. StorageGRID ILM will never delete these items.