CRDB Shows 0 Memory After Creation Despite Specified Size

Redis Enterprise Active-Active (CRDB) databases require memory to be defined per participating instance, and incorrect payload structure or units can result in instances being created with 0 allocated memory. This commonly occurs when using the wrong field (such as a top-level memory_size), incorrect units, or when capacity constraints prevent allocation. This article helps you identify why CRDB memory appears as 0, verify instance configuration, and correct your API payload using Step-by-Step Instructions and Troubleshooting.

Quick Fix Table

Prerequisites

Ensure access and visibility

• Access to Redis Enterprise admin console or API

• Ability to retrieve CRDB and instance configuration (GET requests)

Validate environment readiness

• Sufficient cluster memory and license capacity

• Healthy cluster state with no provisioning failures

Common Causes

Incorrect field used for memory configuration

• CRDB requires memory_limit per instance

• A top-level memory_size field is ignored or defaults to 0

Incorrect units

• memory_limit must be in MB

• Example:

  • 5120 = 5 GB (correct)

  • 5120000 = 5,120 GB (invalid or coerced)

Memory not set per instance

• Each participating cluster must define its own memory_limit

• Missing values can result in 0 allocation

Insufficient capacity or license

• Requested memory exceeds available resources

• Instance may be created but not allocated memory

Invalid shard configuration

• shards_count must be ≥ 1

• Failed deployments may surface as 0 memory

Step-by-Step: Verify CRDB Memory Allocation

1. Check CRDB in admin console

• Open the CRDB

• Inspect each participating instance

• Confirm:

  • Status = healthy

  • Memory ≠ 0

2. Validate via API

• Run a GET request for the CRDB and instances

• Confirm:

  • memory_limit is present per instance

  • Instance status is active/running

3. Check cluster capacity

• Verify available memory on each participating cluster

• Confirm license limits allow requested allocation

Step-by-Step: Fix the Configuration

1. Use the correct field

• Replace top-level memory_size

• Define memory_limit per instance

2. Use correct units (MB)

• 2 GB → 2048

• 5 GB → 5120

• 10 GB → 10240

3. Set shard count

• Ensure shards_count ≥ 1 for each instance

4. Update or recreate the CRDB

• If supported:

  • Use PATCH/PUT to update instance configuration

• Otherwise:

  • Delete and recreate with corrected payload

5. Validate deployment

• Wait for instances to reach healthy state

• Re-check memory allocation in UI/API

Example Payload (Correct Configuration)

{
  "name": "my-crdb",
  "sharding": true,
  "eviction_policy": "allkeys-lru",
  "instances": [
    {
      "name": "my-crdb-us",
      "memory_limit": 5120,
      "shards_count": 1,
      "cluster": {
        "url": "https://cluster-us.example:9443"
      }
    },
    {
      "name": "my-crdb-eu",
      "memory_limit": 5120,
      "shards_count": 1,
      "cluster": {
        "url": "https://cluster-eu.example:9443"
      }
    }
  ]
}

Note: Do not include credentials in logs or shared payloads.

Troubleshooting

Memory still shows 0 after correction

• Confirm all instances include memory_limit

• Verify units are MB, not bytes

Instance not deploying

• Check cluster capacity and license limits

• Review provisioning errors in logs/UI

API accepts payload but ignores memory

• Validate API version and schema expectations

Shard configuration issues

• Ensure shards_count is explicitly set and valid

When to Contact Support

Contact Redis Support if:

• Memory remains 0 after correct configuration and sufficient capacity

• Instances fail to deploy without clear error

• Behavior differs from API documentation

Provide:

• Sanitized request payload

• API endpoint and version

• CRDB and instance IDs

• Timestamps and error messages