使用 Azure 顾问提高应用程序的可靠性Improve the reliability of your application by using Azure Advisor

Azure 顾问可帮助确保并提高业务关键应用程序的连续性。Azure Advisor helps you ensure and improve the continuity of your business-critical applications. 可以从顾问中的顾问仪表板的“可靠性”选项卡上获取可靠性建议。You can get reliability recommendations from Advisor on the Reliability tab of the Advisor dashboard.

检查你的 Check Point 网络虚拟设备映像的版本Check the version of your Check Point network virtual appliance image

顾问可以识别你的虚拟机正在运行的是否是已知会在执行平台服务操作期间失去网络连接的 Check Point 映像版本。Advisor can identify whether your virtual machine is running a version of the Check Point image that has been known to lose network connectivity during platform servicing operations. 顾问建议将帮助你升级到解决了此问题的较新版本的映像。The Advisor recommendation will help you upgrade to a newer version of the image that addresses this problem. 此检查将通过改进网络连接确保业务连续性。This check will ensure business continuity through better network connectivity.

确保应用程序网关容错Ensure application gateway fault tolerance

此建议可确保由应用程序网关提供支持的任务关键型应用程序的业务连续性。This recommendation ensures the business continuity of mission-critical applications that are powered by application gateways. 顾问会识别未配置容错的应用程序网关实例。Advisor identifies application gateway instances that aren't configured for fault tolerance. 然后,它会推荐你可采取的修正操作。It then suggests remediation actions that you can take. 顾问会识别中型或大型单实例应用程序网关,并建议至少再添加一个实例。Advisor identifies medium or large single-instance application gateways and recommends adding at least one more instance. 它还会识别单实例或多实例小型应用程序网关,并建议将它们迁移到中型或大型 SKU。It also identifies single-instance and multiple-instance small application gateways and recommends migrating them to medium or large SKUs. 顾问建议执行这些操作以确保应用程序网关实例配置为满足这些资源的当前 SLA 要求。Advisor recommends these actions to ensure your application gateway instances are configured to satisfy the current SLA requirements for these resources.

防止意外删除虚拟机数据Protect your virtual machine data from accidental deletion

设置虚拟机备份可确保业务关键型数据的可用性,并防止意外删除或损坏。Setting up virtual machine backup ensures the availability of your business-critical data and offers protection against accidental deletion or corruption. 顾问会识别未启用备份的虚拟机,并建议启用备份。Advisor identifies virtual machines where backup isn't enabled and recommends enabling backup.

确保在需要时有权访问 Azure 专家Ensure you have access to Azure experts when you need it

如果你运行的是业务关键型工作负载,那么在需要时有权访问技术支持至关重要。When you're running a business-critical workload, it's important to have access to technical support when you need it. 顾问会识别在其支持计划中不包含技术支持的潜在业务关键型订阅。Advisor identifies potential business-critical subscriptions that don't have technical support included in their support plans. 它会建议升级到包含技术支持的选项。It recommends upgrading to an option that includes technical support.

创建 Azure 服务运行状况警报,以便在 Azure 问题影响你时收到通知Create Azure Service Health alerts to be notified when Azure problems affect you

我们建议设置 Azure 服务运行状况警报,以便在 Azure 服务问题影响你时收到通知。We recommend setting up Azure Service Health alerts so you're notified when Azure service problems affect you. Azure 服务运行状况是一项免费服务,可在你受到 Azure 服务问题影响时提供个性化指导和支持。Azure Service Health is a free service that provides personalized guidance and support when you're affected by an Azure service problem. 顾问会识别未配置警报的订阅,并建议对其进行配置。Advisor identifies subscriptions that don't have alerts configured and recommends configuring them.

配置流量管理器终结点以便进行复原Configure Traffic Manager endpoints for resiliency

如果任何给定的终结点出现故障,则包含多个终结点的 Azure 流量管理器配置文件将体验到更高的可用性。Azure Traffic Manager profiles with more than one endpoint experience higher availability if any given endpoint fails. 将终结点放在不同区域中可进一步提高服务可靠性。Placing endpoints in different regions further improves service reliability. 顾问会识别仅有一个终结点的流量管理器配置文件,并建议在另一个区域中至少再添加一个终结点。Advisor identifies Traffic Manger profiles where there's only one endpoint and recommends adding at least one more endpoint in another region.

如果为邻近路由配置的流量管理器配置文件中的所有终结点都位于同一区域,则来自其他区域的用户可能会遇到连接延迟。If all endpoints in a Traffic Manager profile that's configured for proximity routing are in the same region, users from other regions might experience connection delays. 如果一个区域中的所有终结点都出现故障,则将一个终结点添加或移动到另一个区域会提高整体性能并提供更好的可用性。Adding or moving an endpoint to another region will improve overall performance and provide better availability if all endpoints in one region fail. 顾问会识别为邻近路由配置的流量管理器配置文件(其中所有终结点都位于同一区域中)。Advisor identifies Traffic Manager profiles configured for proximity routing where all the endpoints are in the same region. 它会建议将一个终结点添加或移动到另一个 Azure 区域。It recommends adding or moving an endpoint to another Azure region.

如果为地理路由配置了流量管理器配置文件,则会根据已定义的区域将流量路由到终结点。If a Traffic Manager profile is configured for geographic routing, traffic is routed to endpoints based on defined regions. 如果某个区域出现故障,则不会有任何预定义的故障转移。If a region fails, there's no predefined failover. 如果你的某一终结点的“区域分组”配置为“所有(全球)”,则可避免流量被丢弃,并可提高服务可用性。If you have an endpoint where the Regional Grouping is configured to All (World), you can avoid dropped traffic and improve service availability. 顾问会识别已为地理路由配置的、其中没有任何终结点的“区域分组”配置为“所有(全球)”的流量管理器配置文件。Advisor identifies Traffic Manager profiles configured for geographic routing where there's no endpoint configured to have the Regional Grouping as All (World). 它会建议更改该配置以将一个终结点设为“所有(全球)”。It recommends changing the configuration to make an endpoint All (World).

对 Azure 存储帐户使用软删除可在意外覆盖或删除之后保存和恢复数据Use soft delete on your Azure storage account to save and recover data after accidental overwrite or deletion

为存储帐户启用软删除,以便将删除的 blob 转换为软删除状态,而不是永久删除。Enable soft delete on your storage account so that deleted blobs transition to a soft deleted state instead of being permanently deleted. 覆盖数据时,会生成软删除快照以保存被覆盖数据的状态。When data is overwritten, a soft deleted snapshot is generated to save the state of the overwritten data. 使用软删除可以在意外删除或覆盖时进行恢复。Using soft delete allows you to recover from accidental deletions or overwrites. 顾问会识别未启用软删除的 Azure 存储帐户,并建议你启用软删除。Advisor identifies Azure storage accounts that don't have soft delete enabled and suggests that you enable it.

配置 VPN 网关为主动-主动以进行连接复原Configure your VPN gateway to active-active for connection resiliency

在主动-主动配置中,VPN 网关的两个实例将建立连接到本地 VPN 设备的 S2S VPN 隧道。In active-active configuration, both instances of a VPN gateway establish S2S VPN tunnels to your on-premises VPN device. 当一个网关实例上发生计划内维护事件或计划外事件时,流量将自动切换到其他活动 IPsec 隧道。When a planned maintenance event or unplanned event happens to one gateway instance, traffic is automatically switched to the other active IPsec tunnel. Azure 顾问将识别未配置为主动-主动模式的 VPN 网关,并建议配置它们以提供高可用性。Azure Advisor identifies VPN gateways that aren't configured as active-active and suggests that you configure them for high availability.

使用生产 VPN 网关运行生产工作负荷Use production VPN gateways to run your production workloads

Azure 顾问会检查是否有任何使用基本 SKU 的 VPN 网关,并建议你改用生产 SKU。Azure Advisor checks for any VPN gateways that use a Basic SKU and recommends that you use a production SKU instead. 基本 SKU 设计用于开发和测试。The Basic SKU is designed for development and testing. 生产 SKU 提供:Production SKUs offer:

  • 更多隧道。More tunnels.
  • BGP 支持。BGP support.
  • 主动-主动配置选项。Active-active configuration options.
  • 自定义 IPsec/IKE 策略。Custom Ipsec/IKE policy.
  • 更高的稳定性和可用性。Higher stability and availability.

确保虚拟机容错(暂时禁用)Ensure virtual machine fault tolerance (temporarily disabled)

要为应用程序提供冗余,建议将两个或更多虚拟机分组到一个可用性集中。To provide redundancy for your application, we recommend that you group two or more virtual machines in an availability set. 顾问会识别不属于可用性集的虚拟机,并建议将它们移动到可用性集中。Advisor identifies virtual machines that aren't part of an availability set and recommends moving them into one. 这种配置可以确保在发生计划内或计划外维护期间,至少有一个虚拟机可用,并且它满足 Azure 虚拟机 SLA 要求。This configuration ensures that during either planned or unplanned maintenance, at least one virtual machine is available and meets the Azure virtual machine SLA. 可以选择为虚拟机创建可用性集,或将虚拟机添加到现有可用性集。You can choose to create an availability set for the virtual machine or to add the virtual machine to an existing availability set.

备注

如果选择创建可用性集,则需要至少向其中再添加一个虚拟机。If you choose to create an availability set, you need to add at least one more virtual machine into it. 建议在可用性集中对两个或更多虚拟机进行分组,确保其中一台虚拟机在出现故障期间可用。We recommend that you group two or more virtual machines in an availability set to ensure that at least one machine is available during an outage.

确保可用性集容错(暂时禁用)Ensure availability set fault tolerance (temporarily disabled)

要为应用程序提供冗余,建议将两个或更多虚拟机分组到一个可用性集中。To provide redundancy for your application, we recommend that you group two or more virtual machines in an availability set. 顾问标识包含单个虚拟机的可用性集,并建议向其中添加一个或多个虚拟机。Advisor identifies availability sets that contain a single virtual machine and recommends adding one or more virtual machines to it.  这种配置可以确保在发生计划内或计划外维护期间,至少有一个虚拟机可用,并且它满足 Azure 虚拟机 SLA 要求。This configuration ensures that during either planned or unplanned maintenance, at least one virtual machine is available and meets the Azure virtual machine SLA. 可以选择创建虚拟机,或将现有的虚拟机添加到可用性集。You can choose to create a virtual machine or to add an existing virtual machine to the availability set.

使用托管磁盘提高数据可靠性(暂时禁用)Use managed disks to improve data reliability (temporarily disabled)

如果虚拟机所在的可用性集带有共享存储帐户或存储缩放单元的磁盘,那么这些虚拟机在服务中断期间不可从单个存储规模单元的故障中复原。Virtual machines that are in an availability set with disks that share either storage accounts or storage scale units aren't resilient to failures to single storage scale units during outages. 顾问将识别这些可用性集,并建议迁移到 Azure 托管磁盘。Advisor identifies these availability sets and recommends migrating to Azure managed disks. 此迁移将确保可用性集中的虚拟机的磁盘彼此完全独立,以避免单点故障。This migration will ensure that the disks of the virtual machines in the availability set are sufficiently isolated to avoid a single point of failure.

修复无效的日志警报规则Repair invalid log alert rules

Azure 顾问会检测到在其条件部分指定了无效查询的日志警报规则。Azure Advisor detects log alert rules that have invalid queries specified in their condition section. Azure Monitor 日志警报规则按指定的频率运行查询,并根据结果触发警报。Azure Monitor log alert rules run queries at specified frequency and fire alerts based on the results. 随着时间的推移,查询可能会因所引用资源、表或命令的变化而变得无效。Queries can become invalid over time because of changes in the referenced resources, tables, or commands. 顾问建议对警报查询进行更正,以防止自动禁用规则并确保监视范围。Advisor recommends corrections for alert queries to prevent the rules from being automatically disabled and to ensure monitoring coverage. 有关详细信息,请参阅警报规则故障排除For more information, see Troubleshooting alert rules

在 Azure Cosmos DB 集合上配置“一致”索引模式Configure Consistent indexing mode on your Azure Cosmos DB collection

为 Azure Cosmos DB 容器配置延迟索引模式可能会影响查询结果的新鲜度。Configuring Azure Cosmos DB containers with Lazy indexing mode might affect the freshness of query results. 顾问会检测到以此方式配置的容器,并建议切换到“一致”模式。Advisor detects containers configured this way and recommends switching to Consistent mode. 详细了解 Azure Cosmos DB 中的索引编制策略。Learn more about indexing policies in Azure Cosmos DB.

为 Azure Cosmos DB 容器配置分区键Configure your Azure Cosmos DB containers with a partition key

Azure 顾问将识别即将达到其预配存储配额的 Azure Cosmos DB 非分区集合。Azure Advisor identifies Azure Cosmos DB non-partitioned collections that are approaching their provisioned storage quota. 它会建议将这些集合迁移到具有分区键定义的新集合,使服务能够自动横向扩展它们。It recommends that you migrate these collections to new collections with a partition key definition so that they can be automatically scaled out by the service. 详细了解如何选择分区键。Learn more about choosing a partition key.

将 Azure Cosmos DB .NET SDK 升级到 NuGet 提供的最新版本Upgrade your Azure Cosmos DB .NET SDK to the latest version from NuGet

Azure 顾问会识别使用旧版本的 .NET SDK 的 Azure Cosmos DB 帐户。Azure Advisor identifies Azure Cosmos DB accounts that are using old versions of the .NET SDK. 它会建议升级到 NuGet 提供的最新版本,以获取最新的修补程序、性能改进和功能。It recommends that you upgrade to the latest version from NuGet for the latest fixes, performance improvements, and feature capabilities. 详细了解 Azure Cosmos DB .NET SDK。Learn more about Azure Cosmos DB .NET SDK.

将 Azure Cosmos DB Java SDK 升级到 Maven 提供的最新版本Upgrade your Azure Cosmos DB Java SDK to the latest version from Maven

Azure 顾问会识别使用旧版本的 Java SDK 的 Azure Cosmos DB 帐户。Azure Advisor identifies Azure Cosmos DB accounts that are using old versions of the Java SDK. 它会建议升级到 Maven 提供的最新版本,以获取最新的修补程序、性能改进和功能。It recommends that you upgrade to the latest version from Maven for the latest fixes, performance improvements, and feature capabilities. 详细了解 Azure Cosmos DB Java SDK。Learn more about Azure Cosmos DB Java SDK.

将 Azure Cosmos DB Spark 连接器升级到 Maven 提供的最新版本Upgrade your Azure Cosmos DB Spark connector to the latest version from Maven

Azure 顾问会识别使用旧版本的 Azure Cosmos DB Spark 连接器的 Azure Cosmos DB 帐户。Azure Advisor identifies Azure Cosmos DB accounts that are using old versions of the Azure Cosmos DB Spark connector. 它会建议升级到 Maven 提供的最新版本,以获取最新的修补程序、性能改进和功能。It recommends that you upgrade to the latest version from Maven for the latest fixes, performance improvements, and feature capabilities. 详细了解 Azure Cosmos DB Spark 连接器。Learn more about Azure Cosmos DB Spark connector.

考虑迁移到 HDInsight 4.0 上的 Kafka 2.1Consider moving to Kafka 2.1 on HDInsight 4.0

从 2020 年 7 月 1 日开始,你将无法使用 Azure HDInsight 4.0 上的 Kafka 1.1 创建新的 Kafka 群集。Starting July 1, 2020, you won't be able to create new Kafka clusters by using Kafka 1.1 on Azure HDInsight 4.0. 现有群集将在没有 Microsoft 支持的情况下按原样运行。Existing clusters will run as is without support from Microsoft. 请考虑在 2020 年 6 月 30 日之前迁移到 HDInsight 4.0 上的 Kafka 2.1,避免出现潜在的系统/支持中断。Consider moving to Kafka 2.1 on HDInsight 4.0 by June 30, 2020, to avoid potential system/support interruption.

考虑升级 HDInsight Spark 群集中较旧的 Spark 版本Consider upgrading older Spark versions in HDInsight Spark clusters

从 2020 年 7 月 1 日开始,你将无法使用 HDInsight 3.6 上的 Spark 2.1 或 2.2 创建新的 Spark 群集。Starting July 1, 2020, you won't be able to create new Spark clusters by using Spark 2.1 or 2.2 on HDInsight 3.6. 你将无法使用 HDInsight 4.0 上的 Spark 2.3 创建新的 Spark 群集。You won't be able to create new Spark clusters by using Spark 2.3 on HDInsight 4.0. 现有群集将在没有 Microsoft 支持的情况下按原样运行。Existing clusters will run as is without support from Microsoft.

启用虚拟机复制Enable virtual machine replication

因未启用复制而不能复制到另一区域的虚拟机在出现区域性服务中断时无法复原。Virtual machines that don't have replication enabled to another region aren't resilient to regional outages. 复制虚拟机可在 Azure 区域服务中断期间减少任何不利的业务影响。Replicating virtual machines reduces any adverse business impact during Azure region outages. 顾问会检测到未启用复制的 VM,并建议启用复制。Advisor detects VMs on which replication isn't enabled and recommends enabling it. 启用了复制时,如果发生服务中断,可以使虚拟机快速在远程 Azure 区域中启动。When you enable replication, if there's an outage, you can quickly bring up your virtual machines in a remote Azure region. 详细了解虚拟机复制。Learn more about virtual machine replication.

请勿替代主机名以确保网站的完整性Do not override hostname to ensure website integrity

顾问建议在配置应用程序网关时,尝试避免替代主机名。Advisor recommend to try avoid overriding the hostname when configuring Application Gateway. 如果在应用程序网关前端使用的域不同于访问后端所使用的域,可能会导致 cookie 或重定向 URL 被破坏。Having a different domain on the frontend of Application Gateway than the one which is used to access the backend can potentially lead to cookies or redirect urls being broken. 请注意,并非所有情况都是如此,某种类别的后端(如 REST API 的后端)一般对此不太敏感。Note that this might not be the case in all situations and that certain categories of backends (like REST API's) in general are less sensitive to this. 请确保后端能够应对此情况或更新应用程序网关配置,以便不需要在后端替代主机名。Please make sure the backend is able to deal with this or update the Application Gateway configuration so the hostname does not need to be overwritten towards the backend. 在配合应用服务一起使用时,请将自定义域名附加到 Web 应用,并避免在后端使用 .chinacloudsites.cn 主机名。When used with App Service, attach a custom domain name to the Web App and avoid use of the .chinacloudsites.cn host name towards the backend. 详细了解自定义域Learn more about custom domain.

如何访问顾问中的高可用性建议How to access high availability recommendations in Advisor

  1. 登录 Azure 门户,并打开顾问Sign in to the Azure portal, and then open Advisor.

  2. 在顾问仪表板中,选择“高可用性”选项卡。On the Advisor dashboard, select the High Availability tab.

后续步骤Next steps

有关顾问建议的详细信息,请参阅以下资源:For more information about Advisor recommendations, see: