Configure Windows Server Storage Replica

Source: Internet
Author: User
Tags failover

Storage Replicas (Storage Replica) is a Windows server technology that allows for synchronous replication of volumes between servers or clusters for disaster recovery. It also allows you to use asynchronous replication to create a failover cluster that spans two sites while keeping all nodes synchronized.

Storage replicas support synchronous and asynchronous replication:

    • Synchronous replication maps data in a low-latency network site and a volume that remains consistent during a crash to ensure that the file system-level data loss is zero during a failure.

    • Asynchronous replication maps data between sites outside of the metropolitan area through higher latency network links, but does not guarantee that two sites have identical copies of the data in the event of a failure.

Synchronous replication

Synchronous replication ensures that the application writes data to two locations at a time before the IO is complete. Because this replication requires network and storage investments and has the risk of reducing application performance, it is more appropriate for mission-critical data.

mode illustration step
Synchronous

0 data loss

RPO
650) this.width=650; "Src=" https://i-technet.sec.s-msft.com/zh-cn/windows-server-docs/storage/storage-replica/ Media/storage-replica-overview/storage_sr_synchronousv2.png "alt=" shows the graph of how the storage replica writes data in synchronous replication "style=" border:0px; Vertical-align:middle; "/> 1. Application Writing data
2. Write the log data and copy the data to the remote site
3. Write log data at the remote site
4. Confirm from remote site
5. Confirm that the application writes

T & T1: Data is flushed to the volume, always written to the log
Asynchronous replication

In contrast, asynchronous replication means that when an application writes data, the data is copied to the remote site without an immediate confirmation guarantee. This mode allows for faster response to applications and DR solutions that work geographically.

When an application writes data, the replication engine captures the write and immediately confirms it to the application. The captured data is then copied to the remote location. The remote node processes a copy of the data and delays the return acknowledgment to the source copy. Because replication performance is no longer in the application IO path, the response speed and distance of the remote site are less important factors. If the source data is lost and the destination copy of the data is still in the buffer without leaving the source, there is a risk of data loss.

With a higher version than zero RPO, asynchronous replication is less suitable for HA solutions, such as failover clusters, because they are designed for continuous operations with redundancy and no data loss.

mode illustration step
Asynchronous

Nearly 0 data loss

(depends on many factors)

RPO
650) this.width=650; "Src=" https://i-technet.sec.s-msft.com/zh-cn/windows-server-docs/storage/storage-replica/ Media/storage-replica-overview/storage_sr_asynchronousv2.png "alt=" shows the diagram of how the storage replica writes data in asynchronous Replication "style=" border:0px; Vertical-align:middle; "/> 1. Application Writing data
2. Writing log data
3. Confirm that the application writes
4. Copying data to a remote site
5. Log data is written at the remote site
6. Confirm from remote site

T & T1: Data is flushed to the volume, always written to the log
Application Scenariosstorage replication can be deployed in a stretch cluster set, between cluster-to-cluster configuration and server-to-server configuration.

stretched clusters allow you to configure computers and storage in a single cluster, where some nodes share a set of asymmetric storage, while others share another group, and then synchronize or asynchronously replicate through site-aware. This scenario can take advantage of storage spaces with shared SAS storage, LUNs that are connected to sans and iSCSI. It is managed through the PowerShell and failover Cluster Manager graphical tools and allows automated workload failover.

650) this.width=650; "Src=" https://i-technet.sec.s-msft.com/zh-cn/windows-server-docs/storage/storage-replica/ The Media/storage-replica-overview/storage_sr_stretchcluster.png "alt=" Display uses a storage replica to replicate the storage of the two cluster nodes in new York to new Jersey Two nodes in the diagram "style=" border:0px;vertical-align:middle; "/>

Figure 1: Storage replication using storage replicas in a stretch cluster set

cluster-to-cluster allows replication between two separate clusters, one of which is replicated synchronously or asynchronously with another cluster. This scenario can take advantage of storage spaces direct, storage spaces with shared SAS storage, and LUNs that are connected to sans and iSCSI. It is managed through PowerShell and Azure Site Recovery, and requires manual intervention to failover.

650) this.width=650; "Src=" https://i-technet.sec.s-msft.com/zh-cn/windows-server-docs/storage/storage-replica/ Media/storage-replica-overview/storage_sr_clustertocluster.png "alt=" shows the storage copy of the cluster in Los Angeles as a Las Vegas using a storage replica The different clusters in the diagram "style=" border:0px;vertical-align:middle; "/>

Figure 2: Cluster-to-cluster storage replication using storage replicas

server-to-server allows synchronous and asynchronous replication between two independent servers using storage spaces with shared SAS storage, SAN and ISCSI-connected LUNs, and local drives. It is managed through PowerShell and Server Manager tools, and requires manual intervention to failover.

650) this.width=650; "Src=" https://i-technet.sec.s-msft.com/zh-cn/windows-server-docs/storage/storage-replica/ Media/storage-replica-overview/storage_sr_servertoserver.png "alt=" shows the diagram "style=" that uses the server in build 5 to replicate the servers in build 9 border:0px ; vertical-align:middle; "/>

Figure 3: Server-to-server storage replication using storage replicas

In WindowsServer 2016, the storage replica does not have a GUI interface to configure, so this lab environment is configured with PowerShell:

Required conditions
  • Active Directory Domain Services forest (no need to run Windows Server 2016).

  • Two servers that have Windows server Datacenter Edition installed.

  • Two save sets that use SAS JBOD, Fibre Channel Sans, iSCSI targets, or local scsi/sata storage. Storage needs to include a combination of HDD and SSD media. Set each store to be available only for each server (no shared access).

  • Each save set must allow at least two virtual disks to be created, one for replicated data, and the other for logging.

    The sector size of the physical store on all data disks must be the same. The sector size of the physical store on all log disks must be the same.

  • Each server must have at least one Ethernet/TCP connection for synchronous replication, but preferably RDMA.

  • appropriate firewall and router rules to allow for ICMP, SMB (port 445, and 5445 for SMB passthrough) and Ws-man (Port 5985) bidirectional communication between all nodes.

  • The network between servers has sufficient bandwidth to contain the IO write workload and the average round trip delay of 5 milliseconds (for synchronous replication). Asynchronous replication has no deferred recommendations.

  • The copied storage cannot be located on a drive that contains a Windows operating system folder.



Reference: Https://technet.microsoft.com/zh-cn/windows-server-docs/storage/storage-replica/server-to-server-storage-replication

This article is from "1+1=?" "Blog, be sure to keep this provenance http://lianggj.blog.51cto.com/176264/1905359

Configure Windows Server Storage Replica

Related Article

Contact Us

The content source of this page is from Internet, which doesn't represent Alibaba Cloud's opinion; products and services mentioned on that page don't have any relationship with Alibaba Cloud. If the content of the page makes you feel confusing, please write us an email, we will handle the problem within 5 days after receiving your email.

If you find any instances of plagiarism from the community, please send an email to: info-contact@alibabacloud.com and provide relevant evidence. A staff member will contact you within 5 working days.

A Free Trial That Lets You Build Big!

Start building with 50+ products and up to 12 months usage for Elastic Compute Service

  • Sales Support

    1 on 1 presale consultation

  • After-Sales Support

    24/7 Technical Support 6 Free Tickets per Quarter Faster Response

  • Alibaba Cloud offers highly flexible support services tailored to meet your exact needs.