You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Copy file name to clipboardExpand all lines: articles/cosmos-db/fleet-pools.md
+13-11Lines changed: 13 additions & 11 deletions
Display the source diff
Display the rich diff
Original file line number
Diff line number
Diff line change
@@ -1,12 +1,12 @@
1
1
---
2
-
title: Fleets Pools (Preview)
2
+
title: Fleets Pools
3
3
description: Pools manage throughput in Azure Cosmos DB fleets aimed at optimizing resource allocation for multitenant applications.
4
4
author: deborahc
5
5
ms.author: dech
6
6
ms.service: azure-cosmos-db
7
7
ms.subservice: nosql
8
8
ms.topic: concept-article
9
-
ms.date: 05/07/2025
9
+
ms.date: 02/11/2026
10
10
ai-usage: ai-assisted
11
11
show_latex: true
12
12
appliesto:
@@ -85,20 +85,20 @@ Throughput pooling requires all database accounts in a **fleetspace** to have th
85
85
Here are examples of allowed and unallowed configurations:
86
86
87
87
- Varying regional configurations
88
-
-**Allowed**: Two accounts, both configured as single-region write in two distinct Azure regions.
89
-
-**Allowed**: Two accounts, both configured as multi-region write in two distinct Azure regions.
90
-
-**Not allowed**: One account, with single-region write but distributed to two distinct Azure regions. Another account, with single-region write, in a different single Azure region.
88
+
-**Allowed**: Two accounts, both configured as single-region write (general purpose) in two distinct Azure regions.
89
+
-**Allowed**: Two accounts, both configured as multi-region write (business critical) in two distinct Azure regions.
90
+
-**Not allowed**: One account, with single-region write (general purpose) but distributed to two distinct Azure regions. Another account, with single-region write (general purpose), in a different single Azure region.
91
91
-**Not allowed**: Two accounts, with global distribution configured for two different sets of regions.
92
92
- Varying service tiers
93
93
-**Allowed**: Two accounts, both with multi-region write (business critical) configured in two separate Azure regions.
94
-
-**Not allowed**: Two accounts, one with multi-region write and the other with single-region write.
94
+
-**Not allowed**: Two accounts, one with multi-region write (business critical) and the other with single-region write (general purpose).
95
95
96
96
> [!NOTE]
97
97
> Accounts must have the same underlying configuration of multi-region writes (business critical) or single-region writes (general purpose) to participate in the same pool.
98
98
99
99
## Monitoring consumption
100
100
101
-
In the preview, you can monitor the RU/s the pool scaled to via the `FleetspaceAutoscaledThroughput` metric available at the fleet level.
101
+
You can monitor the RU/s the pool scaled to via the `FleetspaceAutoscaledThroughput` metric available at the fleet level.
102
102
103
103
You can also monitor pooled RU/s consumption at the account level in Azure portal via Azure Monitor following these steps:
104
104
@@ -114,13 +114,15 @@ You can also monitor pooled RU/s consumption at the account level in Azure porta
114
114
115
115
With the pooling feature, the total RU/s available for consumption to each physical partition is still subject to standard [physical partition limits](partitioning.md#physical-partitions). Each **physical partition** has limits on how much extra RU/s it can draw from the pool.
116
116
117
-
In the preview:
117
+
By default:
118
118
119
-
- A physical partition uses up to 5,000 extra RU/s from the pool in addition to its dedicated throughput.
119
+
- A physical partition uses up to 5,000 extra RU/s from the pool in addition to its dedicated throughput. If your application requires a different value, file a support ticket.
120
+
121
+
- A physical partition’s total consumption of dedicated + pool RU/s can't exceed 10,000 RU/s total, even if more RU/s are available to use from the pool.
122
+
123
+
- In summary, the maximum total RU/s a physical partition can consume while using pooling = $\min(5000+currentThroughput, 10000)$.
120
124
121
-
- A physical partition’s total consumption of dedicated + pool RU/s can't exceed 10,000 RU/s total, even if more RU/s are available to use from the pool.
122
125
123
-
- The maximum total RU/s a physical partition can consume while using pooling = $\min(5000+currentThroughput, 10000)$.
124
126
125
127
> [!TIP]
126
128
> You can use the metric `PhysicalPartitionThroughput` in Azure Monitor to determine how many dedicated RU/s are allocated to each physical partition.
0 commit comments