RS: Auto tiering - manage storage engine (#2701)

* manage storage engine DOC-2521 * correct relref * edits * add warning to k8s * Update content/rs/databases/auto-tiering/storage-engine.md Co-authored-by: Alon Magrafta <[email protected]> --------- Co-authored-by: Alon Magrafta <[email protected]>
RedisLabs · Aug 1, 2023 · e4bf327 · e4bf327
1 parent 2a3cce5
commit e4bf327
Show file tree

Hide file tree

Showing 2 changed files with 66 additions and 3 deletions.
diff --git a/content/kubernetes/re-clusters/auto-tiering.md b/content/kubernetes/re-clusters/auto-tiering.md
@@ -32,7 +32,6 @@ Before creating your Redis clusters or databases, these SSDs must be:
 
 For more information on node storage, see [Node persistent and ephemeral storage]({{<relref "/rs/installing-upgrading/install/plan-deployment/persistent-ephemeral-storage">}}).
 
-
 ## Create a Redis Enterprise cluster
 
 To deploy a Redis Enterprise cluster (REC) with flash storage, you'll need to specify the following in the `redisOnFlashSpec` section of your [REC custom resource]({{<relref "/kubernetes/reference/cluster-options.md">}}):
@@ -43,6 +42,8 @@ To deploy a Redis Enterprise cluster (REC) with flash storage, you'll need to sp
 - storage class name (`storageClassName`)
 - minimal flash disk size (`flashDiskSize`)
 
+{{<warning>}}Switching between storage engines requires guidance by Redis Support or your Account Manager.{{</warning>}}
+
 Here is an example of an REC custom resource with these attributes:
 
 ```YAML
@@ -85,5 +86,3 @@ spec:
 {{< note >}}
 This example defines both `memorySize` and `rofRamSize`. When using Auto Tiering, `memorySize` refers to the total combined memory size (RAM + flash) allocated for the database. `rofRamSize` specifies only the RAM capacity for the database. `rofRamSize` must be at least 10% of `memorySize`.
 {{< /note >}}
-
-
diff --git a/content/rs/databases/auto-tiering/storage-engine.md b/content/rs/databases/auto-tiering/storage-engine.md
@@ -0,0 +1,64 @@
+---
+Title: Manage Auto Tiering storage engine
+linkTitle: Manage storage engine
+description: Manage the storage engine used for your database with auto tiering enabled. 
+weight: 100
+alwaysopen: false
+categories: ["RS"]
+aliases: /rs/databases/auto-tiering/storage-engine.md
+---
+
+## Manage the storage engine
+
+Redis Enterprise Auto Tiering supports two storage engines:
+
+* [Speedb](https://www.speedb.io/) (default, recommended)
+* [RocksDB](https://rocksdb.org/)
+
+{{<warning>}}Switching between storage engines requires guidance by Redis Support or your Account Manager.{{</warning>}}
+
+### Change the storage engine
+
+1. Change the cluster level configuration for default storage engine.
+
+  * API:
+
+      ``` sh
+      curl -k -u <username>:<password> -X PUT -H "Content-Type: application/json" -d '{"bigstore_driver":"speedb"}' https://localhost:9443/v1/cluster
+     ```
+
+  * CLI:
+
+      ```sh
+      rladmin cluster config bigstore_driver {speedb | rocksdb}
+      ```
+
+2. Restart the each database on the cluster one by one.
+
+     ```sh
+     rladmin restart db { db:<id> | <name> }
+     ```
+
+{{<note>}} We recommend restarting your database at times with low usage and avoiding peak hours. For databases without persistence enabled, we also recommend using export to backup your database first.{{</note>}}
+
+## Monitor the storage engine
+
+To get the current cluster level default storage engine run:
+
+* Use the `rladmin info cluster` command look for ‘bigstore_driver’.
+
+* Use the REST API:
+
+     ```sh
+     curl -k -u <username>:<password> -X GET -H "Content-Type: application/json" https://localhost:9443/v1/cluster
+     ```
+
+Versions of Redis Enterprise 7.2 and later provide a metric called `bdb_bigstore_shard_count` to help track the shard count per database, filtered by `bdb_id` and by storage engine as shown below:
+
+
+  ```sh
+  bdb_bigstore_shard_count{bdb="1",cluster="mycluster.local",driver="rocksdb"} 1.0
+  bdb_bigstore_shard_count{bdb="1",cluster="mycluster.local",driver="speedb"} 2.0
+  ```
+
+For more about metrics for Redis Enterprise’s integration with Prometheus, see [Prometheus integration]({{<relref "/rs/clusters/monitoring/prometheus-metrics-definitions">}}).