What’s New in SRM and vSphere Replication 8.4
In the last few releases, the Site Recovery Manager and vSphere Replication engineering teams have resolved a number of long-standing customer requests including support for Virtual Volumes (8.3), support for vSphere Replication disk resizing (8.3), Site Recovery Manager as an appliance (8.2), vSphere Replication support for encrypting replication traffic and encrypted virtual machines (8.2) and compatibility with multiple versions of vCenter (8.1).
Given all that they’ve already done, you might think that they’d start slowing down and maybe rest a little bit on their laurels. If you’ve thought that you would be mistaken. Site Recovery Manager 8.4 and vSphere Replication 8.4 have a full array of new features that will delight and satisfy both long-time and newer customers. Let’s take a look at all the cool new stuff that our engineers have cooked up for the SRM and vSphere Replication 8.4 release.
In previous releases of vSphere Replication, when the “Reprotect” workflow was run after a planned migration or failover, vSphere Replication would treat this situation as if replication had just been configured for the first time and a seed file was used. It would run a checksum to compare the disk at the source site (what until the recent planned migration/failover had been the recovery site) to the target site (what until the recent planned/migration had been the protected site) and copy changes wherever they were found. This worked well for the most part, however, it was inefficient because it was usually run soon after the planned migration/failover, and very little of the virtual machine would have changed. And running the checksum process could take a while, especially in the case of larger (>1TB) virtual machines.
With Site Recovery Manager and vSphere Replication 8.4, we’ve enhanced and optimized this process and built more intelligence into how it works. Now after a planned migration, vSphere Replication will automatically start tracking changes that occur to the recovered VMs and when a reprotect workflow is run, only those changes will be replicated to the new recovery site. This has the effect of reducing the amount of time that a reprotect takes from hours/days to seconds/minutes. This time reduction has the benefit of reducing risk. This feature will be explored in more detail in a future post. Note that this capability does require vSphere 7 Update 2 hosts at both sites, in addition to SRM and VR 8.4.
Increased Scale and Improved Performance
vSphere Replication has supported replicating 2,000 VMs per vCenter since version 6.0, or for more than 5 years. Now thanks to many performance tweaks and optimizations from both the vSphere Replication and vSphere engineering teams that number is now 3,000 VMs per vCenter. This is accomplished by replication appliances now supporting up to 300 VMs per appliance (up from 200 VMs per appliance previously). Additional details about these performance optimizations will be highlighted in a future post just know that in addition to greater scale, expect some improvement in replication performance as well.
Almost since its introduction Site Recovery Manager has supported robust reporting on the results of workflows and the steps in recovery plans. In recent versions, both Site Recovery Manager and vSphere Replication have introduced support for exporting and importing their configuration as an XML file. We’ve also introduced the ability to export data from any Datagrid in the UI. Even with all of that, there was the understanding that customers wanted other ways of getting access to detailed information about their recovery plans, protection groups, and protected VMs. That wait is now over. SRM now supports an “Export Global Report” functionality that creates a CSV report with details of all Recovery Plans and Protection groups including protection group type, array-pair/replication servers, storage policy/fault domain, datastore group, VMs, VM IP address, guest OS, Local and Remote devices, hosts, the direction of replication and LUN UUIDs.
User Interface Enhancements
The UI team has added a number of helpful improvements that we’ll review here:
The VM protection workflow has been enhanced to allow for adding missing mappings as part of the protection process. If a VM is being protected and is missing any of its mappings, an option to create those missing mappings will be added to the wizard.
The recovery plan summary page has been improved so that it now displays any warnings and/or errors for vSphere Replication replicated VMs that are a part of the recovery plan along with details where applicable. This makes it easier to understand what is going on in your environment.
Also on the recovery plan summary page, we’ve added functional links to the errors (if there are any). These provide error details without having to open the recovery plan report.
On any screen that includes information about hosts, clusters, or resource pools, the SRM and vSphere Replication interfaces now included details about the hosts, clusters, or resource pools status from vCenter including the current state.
To add clarity when mapping folders a pop-up that provides folder hierarchy has been added. This makes it easier to ensure that the correct folder is selected for mapping.
After completing a failover or planned migration it is important to run the reprotect workflow to reverse replication and reverse protection. This ensures that VMs and their data are protected. A reminder to run the reprotect workflow has been added that appears on the recovery plan summary screen, and as an indicator on the recovery plan tab.
Additionally, when configuring a VM for replication with vSphere Replication in addition to showing disk format and storage policy for the VMs disks, vSphere Replication will now also display the virtual disk controller for each disk.
And lastly, for the enhancements to the UI, there is now a read-only view of the recovery settings wizard for a VM for use when the recovery plan is in a non-editable state.
vSphere Replication Appliance Management Interface
The vSphere Replication appliance has been updated and enhanced in the last few releases and the ongoing improvements have finally reached the management interface. The vSphere Replication appliance now uses the same VMware standard management interface used by Site Recovery Manager, vCenter, and more. It’s now easier than ever to locate settings and complete any management tasks.
Increased Ease of Use
To make the life of an operator easier Site Recovery Manager will now handle a few additional items that previously required manual intervention. First, Site Recovery Manager now provides the capability to automatically select placeholder datastores at both the protected and recovery sites. It has intelligence built into it to ensure that the datastores it selects meet all the requirements for use as placeholder datastores. This capability can be turned off (it is enabled by default) using an advanced setting.
Second, VMs that will be protected, and have ISOs attached will rarely need those ISOs attached when they are recovered. Site Recovery Manager now provides an advanced setting that will exclude CD-ROM and floppy drives from protection or recovery. In 8.4 this setting is enabled by default.
Enhanced Pairing Support
To make it easier to upgrade, as it isn’t always possible for both sites at the same time, both SRM and vSphere Replication now support running different versions (n+1 max) at the protected and recovery sites. This is recommended primarily for during an upgrade process and there are a few things to be aware of:
- The pair will function with the features and capabilities of the older version.
- The versions of SRM and vSphere Replication must be the same within each site (eg. SRM 8.4 and vSphere Replication 8.3 at either the protected or recovery site would not be supported).
- This is a separate capability from the support for different versions of vCenter at the protected and recovery site. That feature is not limited to upgrade periods.
Engineering has continued to expand and improve upon the SRM API by adding the following new capabilities:
- View number and type of licenses
- Create empty protection groups
- Create/edit folders and move recovery plans and protection groups
- Configure array managers
- Manage replicated array pairs
- Manual per VM protection/inventory mapping
- Add/Remove/Get Placeholder Datastore(s)
- Pair/Reconfigure/Break Site Recovery Manager services
- Add VM dependencies
Additionally, the vRealize Orchestrator plug-in for SRM has added:
- Add Inventory Mappings per VM
- Configure Placeholder Datastore
- New workflow to create Protection Group and Recovery Plan folders and move Protection Groups and Recovery Plans from one folder into another
And the vRealize Orchestrator plug-in for vSphere Replication improved configure replication workflow. It has enabled configuration and reconfiguration of datastore per disk and storage policy per disk.
vRealize Operations Management Packs
Last and certainly not least, new capabilities have been added to both of the management packs for vRealize Operations. Additionally, the updated management packs now support vRealize Operations Manager 8.3. The new details that are available to report on are:
- Licensing details
- Array Manager Objects
- Errors, Alerts, and Alarms
- Recovery History
- And for vSphere Replication, MPIT visibility
As you can see there is a lot to this release and there should be something here for every user of SRM and vSphere Replication 8.4. Look for the follow-up post on the reprotect optimization and tell us what you think in the comments or even better reach out on Twitter.