Skip to content

Commit 646d6ab

Browse files
authored
Merge pull request #177885 from deborahc/dech-cosmos-partitioning-docs
Adding info on how splits get paused during add region and failover operations
2 parents 8c76448 + dff7b74 commit 646d6ab

File tree

5 files changed

+18
-4
lines changed

5 files changed

+18
-4
lines changed

articles/cosmos-db/how-to-manage-database-account.md

Lines changed: 4 additions & 1 deletion
Original file line numberDiff line numberDiff line change
@@ -25,7 +25,7 @@ This article describes how to manage various tasks on an Azure Cosmos account us
2525
## Add/remove regions from your database account
2626

2727
> [!TIP]
28-
> When a new region is added, all data must be fully replicated and committed into the new region before the region is marked as available. The amount of time this operation takes will depend upon how much data is stored within the account.
28+
> When a new region is added, all data must be fully replicated and committed into the new region before the region is marked as available. The amount of time this operation takes will depend upon how much data is stored within the account. If an [asynchronous throughput scaling operation](scaling-provisioned-throughput-best-practices.md#background-on-scaling-rus) is in progress, the throughput scale-up operation will be paused and will resume automatically when the add/remove region operation is complete.
2929
3030
1. Sign in to [Azure portal](https://portal.azure.com).
3131

@@ -91,6 +91,9 @@ After a Cosmos account is configured for automatic failover, the failover priori
9191
> [!IMPORTANT]
9292
> The Azure Cosmos account must be configured for manual failover for this operation to succeed.
9393
94+
> [!NOTE]
95+
> If you perform a manual failover operation while an [asynchronous throughput scaling operation](scaling-provisioned-throughput-best-practices.md#background-on-scaling-rus) is in progress, the throughput scale-up operation will be paused. It will resume automatically when the failover operation is complete.
96+
9497
1. Go to your Azure Cosmos account, and open the **Replicate data globally** menu.
9598

9699
2. At the top of the menu, select **Manual Failover**.

articles/cosmos-db/how-to-move-regions.md

Lines changed: 3 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -36,6 +36,9 @@ Azure Cosmos DB supports data replication natively, so moving data from one regi
3636

3737
To remove a region from an Azure Cosmos DB account, see [Add/remove regions from your Azure Cosmos DB account](how-to-manage-database-account.md#addremove-regions-from-your-database-account).
3838

39+
> [!NOTE]
40+
> If you perform a failover operation or add/remove a new region while an [asynchronous throughput scaling operation](scaling-provisioned-throughput-best-practices.md#background-on-scaling-rus) is in progress, the throughput scale-up operation will be paused. It will resume automatically when the failover or add/remove region operation is complete.
41+
3942
## Migrate Azure Cosmos DB account metadata
4043

4144
Azure Cosmos DB does not natively support migrating account metadata from one region to another. To migrate both the account metadata and customer data from one region to another, you must create a new account in the desired region and then copy the data manually.

articles/cosmos-db/scaling-provisioned-throughput-best-practices.md

Lines changed: 3 additions & 1 deletion
Original file line numberDiff line numberDiff line change
@@ -30,10 +30,12 @@ When you send a request to increase the RU/s of your database or container, depe
3030
- As a result, the operation can take some time to complete, typically 4-6 hours.
3131
Each physical partition can support a maximum of 10,000 RU/s (applies to all APIs) of throughput and 50 GB of storage (applies to all APIs, except Cassandra, which has 30 GB of storage).
3232

33+
> [!NOTE]
34+
> If you perform a [manual region failover operation](how-to-manage-database-account.md#manual-failover) or [add/remove a new region](how-to-manage-database-account.md#addremove-regions-from-your-database-account) while an asynchronous scale-up operation is in progress, the throughput scale-up operation will be paused. It will resume automatically when the failover or add/remove region operation is complete.
3335
- **Instant scale-down**
3436
- For scale down operation Azure Cosmos DB doesn’t need to split or add new partitions.
3537
- As a result, the operation completes immediately and the RU/s are available for use,
36-
- Key result of this operation is that RUs per partition range will be reduced.
38+
- Key result of this operation is that RUs per physical partition will be reduced.
3739

3840
## How to scale-up RU/s without changing partition layout
3941

articles/cosmos-db/sql/manage-with-cli.md

Lines changed: 4 additions & 1 deletion
Original file line numberDiff line numberDiff line change
@@ -66,7 +66,7 @@ Create an Azure Cosmos account with two regions, add a region, and remove a regi
6666
> [!NOTE]
6767
> This command allows you to add and remove regions but does not allow you to modify failover priorities or trigger a manual failover. See [Set failover priority](#set-failover-priority) and [Trigger manual failover](#trigger-manual-failover).
6868
> [!TIP]
69-
> When a new region is added, all data must be fully replicated and committed into the new region before the region is marked as available. The amount of time this operation takes will depend upon how much data is stored within the account.
69+
> When a new region is added, all data must be fully replicated and committed into the new region before the region is marked as available. The amount of time this operation takes will depend upon how much data is stored within the account. If an [asynchronous throughput scaling operation](../scaling-provisioned-throughput-best-practices.md#background-on-scaling-rus) is in progress, the throughput scale-up operation will be paused and will resume automatically when the add/remove region operation is complete.
7070
7171
```azurecli-interactive
7272
resourceGroupName='myResourceGroup'
@@ -139,6 +139,9 @@ az cosmosdb update --ids $accountId --enable-automatic-failover true
139139
> [!CAUTION]
140140
> Changing region with priority = 0 will trigger a manual failover for an Azure Cosmos account. Any other priority change will not trigger a failover.
141141
142+
> [!NOTE]
143+
> If you perform a manual failover operation while an [asynchronous throughput scaling operation](../scaling-provisioned-throughput-best-practices.md#background-on-scaling-rus) is in progress, the throughput scale-up operation will be paused. It will resume automatically when the failover operation is complete.
144+
142145
```azurecli-interactive
143146
# Assume region order is initially 'West US 2=0' 'East US 2=1' 'South Central US=2' for account
144147
resourceGroupName='myResourceGroup'

articles/cosmos-db/sql/manage-with-powershell.md

Lines changed: 4 additions & 1 deletion
Original file line numberDiff line numberDiff line change
@@ -116,7 +116,7 @@ This command allows you to update your Azure Cosmos DB database account properti
116116
> [!NOTE]
117117
> This command allows you to add and remove regions but does not allow you to modify failover priorities or trigger a manual failover. See [Modify failover priority](#modify-failover-priority) and [Trigger manual failover](#trigger-manual-failover).
118118
> [!TIP]
119-
> When a new region is added, all data must be fully replicated and committed into the new region before the region is marked as available. The amount of time this operation takes will depend upon how much data is stored within the account.
119+
> When a new region is added, all data must be fully replicated and committed into the new region before the region is marked as available. The amount of time this operation takes will depend upon how much data is stored within the account. If an [asynchronous throughput scaling operation](../scaling-provisioned-throughput-best-practices.md#background-on-scaling-rus) is in progress, the throughput scale-up operation will be paused and will resume automatically when the add/remove region operation is complete.
120120
121121
```azurepowershell-interactive
122122
# Create account with two regions
@@ -319,6 +319,9 @@ For the example below, assume the account has a current failover priority of `We
319319
> [!CAUTION]
320320
> Changing `locationName` for `failoverPriority=0` will trigger a manual failover for an Azure Cosmos account. Any other priority change will not trigger a failover.
321321
322+
> [!NOTE]
323+
> If you perform a manual failover operation while an [asynchronous throughput scaling operation](../scaling-provisioned-throughput-best-practices.md#background-on-scaling-rus) is in progress, the throughput scale-up operation will be paused. It will resume automatically when the failover operation is complete.
324+
322325
```azurepowershell-interactive
323326
$resourceGroupName = "myResourceGroup"
324327
$accountName = "mycosmosaccount"

0 commit comments

Comments
 (0)