你当前正在访问 Microsoft Azure Global Edition 技术文档网站。 如果需要访问由世纪互联运营的 Microsoft Azure 中国技术文档网站,请访问 https://docs.azure.cn。
将每个 Azure Cosmos DB 资源的吞吐量从标准转换为自动缩放
适用对象: NoSQL MongoDB Cassandra Gremlin 表
本文中的脚本演示了如何将订阅中的每个资源从标准预配吞吐量转换为自动缩放。
许多客户在开发新应用程序时,会首先选择标准预配吞吐量。 然而,标准吞吐量最适合用于具有持续吞吐量需求的工作负载。 大多数工作负载都是可变的。 这意味着自动缩放的使用成本通常更低。 在可能有数十或数百个要迁移的资源的情况下,这可能会很繁琐且耗时。 此脚本用于在单个步骤中迁移所有资源。
如果没有 Azure 订阅,请在开始之前创建一个 Azure 免费帐户。
先决条件
在 Azure Cloud Shell 中使用 Bash 环境。 有关详细信息,请参阅 Azure Cloud Shell 中的 Bash 快速入门。
如需在本地运行 CLI 参考命令,请安装 Azure CLI。 如果在 Windows 或 macOS 上运行,请考虑在 Docker 容器中运行 Azure CLI。 有关详细信息,请参阅如何在 Docker 容器中运行 Azure CLI。
如果使用的是本地安装,请使用 az login 命令登录到 Azure CLI。 若要完成身份验证过程,请遵循终端中显示的步骤。 有关其他登录选项,请参阅使用 Azure CLI 登录。
出现提示时,请在首次使用时安装 Azure CLI 扩展。 有关扩展详细信息,请参阅使用 Azure CLI 的扩展。
运行 az version 以查找安装的版本和依赖库。 若要升级到最新版本,请运行 az upgrade。
- 本文需要 Azure CLI 版本 2.9.1 或更高版本。 如果使用 Azure Cloud Shell,则最新版本已安装。
示例脚本
启动 Azure Cloud Shell
Azure Cloud Shell 是免费的交互式 shell,可以使用它运行本文中的步骤。 它预安装有常用 Azure 工具并将其配置与帐户一起使用。
若要打开 Cloud Shell,只需要从代码块的右上角选择“试一试”。 也可以通过转到 https://shell.azure.com 在单独的浏览器标签页中启动 Cloud Shell。
当 Cloud Shell 打开时,请验证是否为环境选择了“Bash”。 后续会话将在 Bash 环境中使用 Azure CLI,选择“复制”以复制代码块,将其粘贴到 Cloud Shell 中,然后按 Enter 来运行它。
登录 Azure
Cloud Shell 会在登录时使用的初始帐户下自动进行身份验证。 使用以下脚本通过其他订阅登录,将 <Subscription ID>
替换为 Azure 订阅 ID。 如果没有 Azure 订阅,请在开始之前创建一个 Azure 免费帐户。
subscription="<subscriptionId>" # add subscription here
az account set -s $subscription # ...or use 'az login'
运行脚本
#!/bin/bash
# Passed validation in Cloud Shell on 7/25/2024
# <FullScript>
# Azure Cosmos DB users can migrate from standard provisioned to autoscale
# throughput and back again. Most users can benefit from autoscale throughput
# to save on costs and avoid over-provisioning. This script migrates all
# provisioned throughput to autoscale throughput for all Cosmos DB accounts
# in the current subscription for NoSQL, MongoDB, Cassandra, Gremlin, and Table
# database and container level resources in the accounts.
# These can remain commented out if running in Azure Cloud Shell
#az login
#az account set -s {your subscription id}
throughtput=0
# Get the list of resource groups in the current subscription
resourceGroups=$(az group list --query "[].name" -o tsv)
# Loop through every resource group in the subscription
for resourceGroup in $resourceGroups; do
echo "Processing resource group: $resourceGroup"
# Get the list of Cosmos DB accounts in this resource group
accounts=$(az cosmosdb list -g $resourceGroup --query "[].name" -o tsv)
# Loop through every Cosmos DB account in the resource group
for account in $accounts; do
echo "Processing account: $account"
# Get the list of SQL databases in this account
databases=$(az cosmosdb sql database list -g $resourceGroup -a $account --query "[].name" -o tsv)
# Loop through SQL databases in the account
for database in $databases; do
throughput=$(az cosmosdb sql database throughput show -g $resourceGroup -a $account -n $database --query resource.throughput -o tsv)
if [ $throughput -gt 0 ]; then
echo "$database has throughput, attempting to migrate to autoscale"
# Migrate the database to autoscale throughput
az cosmosdb sql database throughput migrate -g $resourceGroup -a $account -n $database -t "autoscale"
if [ $? -eq 0 ]; then
echo "Successfully migrated throughput for database: $database in Cosmos DB account $account"
fi
else
echo "$database does not have throughput"
fi
# Loop through SQL containers in the database
containers=$(az cosmosdb sql container list -g $resourceGroup -a $account -d $database --query "[].name" -o tsv)
for container in $containers; do
throughput=$(az cosmosdb sql container throughput show -g $resourceGroup -a $account -d $database -n $container --query resource.throughput -o tsv)
if [ $throughput -gt 0 ]; then
echo "$container has throughput, attempting to migrate to autoscale"
# Migrate the container to autoscale throughput
az cosmosdb sql container throughput migrate -g $resourceGroup -a $account -d $database -n $container -t "autoscale"
if [ $? -eq 0 ]; then
echo "Successfully migrated throughput for container: $container in Cosmos DB account $account and database $database"
fi
else
echo "$container does not have throughput"
fi
done
done
# Get the list of MongoDB databases in this account
databases=$(az cosmosdb mongodb database list -g $resourceGroup -a $account --query "[].name" -o tsv)
# Loop through MongoDB databases in the account
for database in $databases; do
throughput=$(az cosmosdb mongodb database throughput show -g $resourceGroup -a $account -n $database --query resource.throughput -o tsv)
if [ $throughput -gt 0 ]; then
echo "$database has throughput, attempting to migrate to autoscale"
# Migrate the database to autoscale throughput
az cosmosdb mongodb database throughput migrate -g $resourceGroup -a $account -n $database -t "autoscale"
if [ $? -eq 0 ]; then
echo "Successfully migrated throughput for database: $database in Cosmos DB account $account"
fi
else
echo "$database does not have throughput"
fi
# Loop through MongoDB collections in the database
collections=$(az cosmosdb mongodb collection list -g $resourceGroup -a $account -d $database --query "[].name" -o tsv)
for collection in $collections; do
throughput=$(az cosmosdb mongodb collection throughput show -g $resourceGroup -a $account -d $database -n $collection --query resource.throughput -o tsv)
if [ $throughput -gt 0 ]; then
echo "$collection has throughput, attempting to migrate to autoscale"
# Migrate the collection to autoscale throughput
az cosmosdb mongodb collection throughput migrate -g $resourceGroup -a $account -d $database -n $collection -t "autoscale"
if [ $? -eq 0 ]; then
echo "Successfully migrated throughput for collection: $collection in Cosmos DB account $account and database $database"
fi
else
echo "$collection does not have throughput"
fi
done
done
# Get the list of Cassandra keyspaces in this account
keyspaces=$(az cosmosdb cassandra keyspace list -g $resourceGroup -a $account --query "[].name" -o tsv)
# Loop through Cassandra keyspaces in the account
for keyspace in $keyspaces; do
throughput=$(az cosmosdb cassandra keyspace throughput show -g $resourceGroup -a $account -n $keyspace --query resource.throughput -o tsv)
if [ $throughput -gt 0 ]; then
echo "$keyspace has throughput, attempting to migrate to autoscale"
# Migrate the keyspace to autoscale throughput
az cosmosdb cassandra keyspace throughput migrate -g $resourceGroup -a $account -n $keyspace -t "autoscale"
if [ $? -eq 0 ]; then
echo "Successfully migrated throughput for keyspace: $keyspace in Cosmos DB account $account"
fi
else
echo "$keyspace does not have throughput"
fi
# Loop through Cassandra tables in the keyspace
tables=$(az cosmosdb cassandra table list -g $resourceGroup -a $account -k $keyspace --query "[].name" -o tsv)
for table in $tables; do
throughput=$(az cosmosdb cassandra table throughput show -g $resourceGroup -a $account -k $keyspace -n $table --query resource.throughput -o tsv)
if [ $throughput -gt 0 ]; then
echo "$table has throughput, attempting to migrate to autoscale"
# Migrate the table to autoscale throughput
az cosmosdb cassandra table throughput migrate -g $resourceGroup -a $account -k $keyspace -n $table -t "autoscale"
if [ $? -eq 0 ]; then
echo "Successfully migrated throughput for table: $table in Cosmos DB account $account and keyspace $keyspace"
fi
else
echo "$table does not have throughput"
fi
done
done
# Get the list of Gremlin databases in this account
databases=$(az cosmosdb gremlin database list -g $resourceGroup -a $account --query "[].name" -o tsv)
# Loop through Gremlin databases in the account
for database in $databases; do
throughput=$(az cosmosdb gremlin database throughput show -g $resourceGroup -a $account -n $database --query resource.throughput -o tsv)
if [ $throughput -gt 0 ]; then
echo "$database has throughput, attempting to migrate to autoscale"
# Migrate the database to autoscale throughput
az cosmosdb gremlin database throughput migrate -g $resourceGroup -a $account -n $database -t "autoscale"
if [ $? -eq 0 ]; then
echo "Successfully migrated throughput for database: $database in Cosmos DB account $account"
fi
else
echo "$database does not have throughput"
fi
# Loop through Gremlin graphs in the database
graphs=$(az cosmosdb gremlin graph list -g $resourceGroup -a $account -d $database --query "[].name" -o tsv)
for graph in $graphs; do
throughput=$(az cosmosdb gremlin graph throughput show -g $resourceGroup -a $account -d $database -n $graph --query resource.throughput -o tsv)
if [ $throughput -gt 0 ]; then
echo "$graph has throughput, attempting to migrate to autoscale"
# Migrate the graph to autoscale throughput
az cosmosdb gremlin graph throughput migrate -g $resourceGroup -a $account -d $database -n $graph -t "autoscale"
if [ $? -eq 0 ]; then
echo "Successfully migrated throughput for graph: $graph in Cosmos DB account $account and database $database"
fi
else
echo "$graph does not have throughput"
fi
done
done
# Get the list of Table databases in this account
tables=$(az cosmosdb table list -g $resourceGroup -a $account --query "[].name" -o tsv)
# Loop through Table databases in the account
for table in $tables; do
throughput=$(az cosmosdb table throughput show -g $resourceGroup -a $account -n $table --query resource.throughput -o tsv)
if [ $throughput -gt 0 ]; then
echo "$table has throughput, attempting to migrate to autoscale"
# Migrate the table to autoscale throughput
az cosmosdb table throughput migrate -g $resourceGroup -a $account -n $table -t "autoscale"
if [ $? -eq 0 ]; then
echo "Successfully migrated throughput for table: $table in Cosmos DB account $account"
fi
else
echo "$table does not have throughput"
fi
done
done
done
echo "All Done! Enjoy your new autoscale throughput Cosmos DB accounts!"
# </FullScript>
示例参考
此脚本使用以下命令。 表中的每条命令均链接到特定于命令的文档。
后续步骤
有关 Azure Cosmos DB CLI 的详细信息,请参阅 Azure Cosmos DB CLI 文档。
有关特定 API 的 Azure CLI 示例,请参阅: