Posted by
Ken Werneburg
Tech Marketing
Twitter @vmKen
Today at VMworld we announced a bunch of very exciting technologies, for this article I’ll be talking first about Site Recovery Manager 5.1, and then about vSphere Replication as a standalone feature of the vSphere platform.
Protection of your systems is a critical aspect of running a virtual infrastructure, and with these announcements (and that of vSphere Data Protection) we’ve really rounded out the business continuity functions of vSphere.
This is a fairly small release with some great features that continue to deliver on the changes we introduced with 5.0. At a high level, the changes are:
- Improved VSS integration for quiescent applications with vSphere Replication
- Improved storage handling for quicker and more consistent responsiveness and behaviour
- Forced recovery with vSphere Replication
- Reprotect and fallback with vSphere Replication
- A move to a 64 bit process
- Support for Essentials Plus environments.
Now a little more detail about a few of these items.
VSS
VMware Tools has the ability to issue commands to the operating system such as to set up VSS snapshots. With 5.1 we have the ability to do a little more than we have in the past, and ask the OS to flush application writers as well as make the OS itself quiescent. This means for things like databases, messaging platforms, and other applications that have VSS writers, we can ensure a higher level of application recoverability. When using vSphere Replication we can flush all the writers for the apps and the OS ensuring data consistency for the image used for recovery.
It’s completely transparent to the OS, and a simple drop down that is chosen when setting up replication for a Windows VM.
Forced Recovery
In SRM 5.0.1 we introduced the forced failover ability: If your primary site is down or responding inconsistently sometimes we might have timeouts and errors waiting for results. This option for failover ensures that only recovery-side operations take place, and we don’t timeout waiting for commands to return from the protected site. This was, at the time, only possible to use with array replication. With 5.1 it is now supported for vSphere Replication as well.
Reprotect and Failback
We can now, after failing over, simply click the “reprotect” button and the environment that has moved to the secondary site will be fully protected back to the original site, irrespective of type of replication you’re using. Reprotect for vSphere Replication is fantastic – it’ll use the existing policies of replication, protection groups, do a full sync back to the primary, and you are then ready to recover or migrate back to the primary location!
Essentials Plus support
One of the most numerous requests we’ve received over the years is to make SRM more accessible to the small and midsize business market. This step to make SRM compatible with Essentials Plus makes disaster recovery more accessible than ever for the SMB customers who have as much need for business continuity as every other customer!
Now on to vSphere Replication
vSphere Replication was introduced with SRM 5.0 as a means of protecting VM data using our in-hypervisor software based replication. It was part of SRM 5.0, and continues to be, carrying forward, but now we are offering the ability to use this technology in a new fashion.
Today’s announcement about vSphere Replication is a big one: We have decoupled it from SRM and released it as an available feature of every vSphere license from Essentials Plus through Enterprise Plus.
Every customer can now protect their environment, using vSphere Replication as a fundamental feature of the protection of your environment, just like HA.
VR does not include all the orchestration, testing, reporting and enterprise-class DR functions of SRM, but allows for individual VM protection and recovery within or across clusters. For many customers this type of protection is critical and has been difficult to attain short of buying into a full multisite DR solution with SRM. Now most of our customers can take advantage of virtual machine protection and recovery with vSphere Replication.
Check out an introduction to vSphere Replication at http://www.vmware.com/files/pdf/techpaper/Introduction-to-vSphere-Replication.pdf
Hi Ken
The failback option in sphere replicator is always full sync or could it be incremental in case the data is still available on the old Protected site?
Thanks
I believe vSphere Replication will use the existing data as a seed for the replication so only the differences will have to be synchronized back.
Hi Satinder, doing a reprotect it will always do a full sync, meaning it will fully determine what blocks are out of sync with each other. So it will *calculate* a full sync, but the blocks that get shipped will only be the differential, presuming the data is still there.
Hi Ken,
“VR does not include all the orchestration, testing…”
Is there any way to test the snapshot before failing over?
Does it mean I’ll need to failover without checking the snapshot first? It means that corrupted snapshot on the production site will result also with a corrupted snapshot on the secondary site?
Hi Vered,
There is no way to test the failover if you are not using SRM. The snapshot is protected from writing garbage at the VMDK level by ensuring that only the fully replicated data set gets committed, but you are correct that if what is replicated is ‘bad OS data’ then that data gets written to the VM too.
Keep in mind it’s not a backup solution, so having backup is still very important for those scenarios!
When does SRM 5.1 get released?
Well, historically we’ve released the products pretty quickly after the announcements.
Pingback: Storage Changes in VMware vSphere 5.1 – @SFoskett – Stephen Foskett, Pack Rat
SRM and vSphere replication are both great products but seems more targeted towards large companies or service providers. What about SMBs who might be hosted in colos but cannot afford the huge pipes either for SAN array replication or data replication for maintaining a DR site? We have two locations on the West and the East coast, hosted by different vendors and would love to use SRM or vSphere replication instead of database log shipping to keep the sites in sync but unfortunately don’t have a cost effective solution. I’m sure there are other companies on the same boat. Hopefully VMware will take a look at this segment also.
Pingback: Welcome to vSphere-land! » vSphere 5.1 Link-O-Rama
Hi Ken,
vSphere Replication administration guide state that:
“…The available quiescing types are determined by the virtual machine’s operating system. Microsoft Volume
Shadow Copy Service (VSS) quiescing is supported for Windows virtual machines running Windows Server 2003, XP, or newer…” and than also “… vSphere Replication supports file-system level quiescing for Windows 7 and Windows 2008 and application-level quiescing for Windows Server 2008 and Windows Server 8 operating systems…” .
So: what Kind of VSS quiescing is supported foe W2003 if not the file-leve and application-level one ?
Thanks very much.
Pier
It will still do the VSS call, but not attempt to flush application writers so you will get a good OS/FS quiescence, at a minimum.
Pingback: Complete guide to all the essential vSphere 5.1 links | Wintual
Can you give us a quick pointer on the key advantages of using SRM over vSphere Replication as it stands today?
thanks!
The differences are plenty: Automation, test ability, reporting, orchestration, scripting, OS customization, multi-VM recovery, an API, and more!
Hi Ken,
am I right that with vSphere Replication 5.1 you only need 1 Appliance for replication per site and not 1 VRMS and at least 1 VRS, as it was in SRM 5.0 ??
Would be much better, because implementing Replication with SRM 5.0 was not that fun.
br
Stefan
You are correct Stefan, one appliance per site, no need for VRMS + VRS, etc.
You can still deploy further VRS appliances, still, if you decide you need them though!
Pingback: VMworld 2012 Recap — latoga labs
Hi Ken,
I have couple of queries for SRM and vSphere 5.1 Replication
1.) Don’t vSphere 5.1 Replication support VMs that use storage DRS.
2.) Which is a better solution viz Storage based replication or vsphere 5 replication
3.) If we have an VDI enviornment (Citrix) on vSphere can we protect the virtual desktops
4.) I noticed that vSphere replication support upto 500 VMs, is there any cap on ESX host.
BR
1) No storage DRS or storage vMotion with vSphere Replication yet – that’s in the works though!
2) This is a hard one to say – VR has a few limitations (15 minute most-aggressive RPO, no powered off VM, template, etc. replication, no consistency groups, etc.) but it allows per-VM replication policies. Storage replication is extremely powerful but usually tied to replicating a full disk group/LUN/etc. at its most granular level. They are both good, and often we will talk about using both, with VR as an augmentation to ABR.
3) You can replicate a VDI environment, and jump through a few hoops to recover them with SRM, but usually it’s better to build a site-resilient VDI infrastructure with a site load balancer in place.
4) No cap on hosts – each host has its own VR agent within the kernel. Each VR appliance is connected to a VC so the VC capability will limit the number of hosts before VR does.
Hello Ken
I’ve been testing vSphere Replication and I haven’t seen the ability to reprotect a VM other than to reconfigure replication. I was wondering if there was a reprotect option in vSphere Replication after a VM is recovered? Or is that only available when using SRM?
Thank you
James
You are correct that automated reprotect is only available through SRM. With VR on its own you’ll need to manually configure the protection again in the opposite direction. If the original VM is still there keep in mind you can use that as a seed for replication!
Pingback: VMTN Blog: VMware Communities Roundtable Podcast – Show Notes #206 – What’s New in SRM 5.1 | Strategic HRStrategic HR
Howdy! This is kind of off topic but I need some guidance from an established blog.
Is it difficult to set up your own blog? I’m not very techincal but I can figure things out pretty fast. I’m thinking about
creating my own but I’m not sure where to start. Do you have any tips or suggestions? Cheers
Hi, after reading this awesome paragraph i am also cheerful to share my familiarity here with mates.
Hi, if I protect a vm with VR and at time x I power off the vm, if I recover it on the secondary site at time x+15 minutes I will recover the vm state at time x?
If you power it off it will then stop driving IO and the last replication may or may not include the final “x” bits of the poweroff.
If you try to recover the VM and the source VM is still available (even powered off) then VR will grab the most recent data (x+15).
So it depends on the state of the VM being recovered. If it’s unavailable, then you get it at point X. If it’s still available for a final synch then you get it at X+15.
Hi there this is kinda of off topic but I was wanting to know
if blogs use WYSIWYG editors or if you have to manually code
with HTML. I’m starting a blog soon but have no coding expertise so I wanted to get guidance from someone with experience. Any help would be enormously appreciated!
Wonderful goods from you, man. I have understand your stuff prior to and you’re simply extremely magnificent. I actually like what you’ve acquired right here,
certainly like what you’re stating and the best way in which you are saying it. You make it entertaining and you continue to care for to stay it smart. I cant wait to learn far more from you. That is actually a great website.
I blog frequently and I genuinely thank you for your information.
This great article has really peaked my interest. I will take a note of your
website and keep checking for new details about once a week.
I opted in for your Feed too.
Hi ken, do we need to purchase additional license for VR appliance or it is bundled with acceleration kit, one more question in 5.1 replication takes place over management port or ” vmkernal port which is marked for REPLICATION”
Hi Kamlesh,
VR is included with every edition of vSphere from Essentials Plus up.
So basically if you have the free ESXi vSphere Hypervisor, you can’t use VR as it requires a vCenter instance, and if you have vSphere Essentials, you can’t use VR by license restriction. Anything other than that includes a license to use VR for free.
With regards to the second question, it takes place over the management vmkernel port that has correct routing. The tag for replication is strictly visual, it does not change the traffic flow.
In order to push replication traffic over a specific vmkernel port group you will need to add a static route to the host to use that port group.
Hi Ken,
I need your recommendations on below points for designing a DR solution using SRM
1.) How much compute capacity(CPU,memory/disk) we should allocate for SRM VM at each site. Is there any recommendation guide for sizing of SRM w.r.t no of hosts or VM protected?
2.) For Testing out DR plans how much additional storage capacity should we keep in buffer at fail over site as I understand that SRM makes snapshots of virtual machines during DR plan testing. Can we also keep that additional capacity on NL disks if my primary data sits on high performance FC or SAS disks?
3.) As a general best practice shall we keep the IP schema same as that of production VMs or should they be different?
BR
1) Totally dependent on the number and sizing of VMs you intend to fail over, and your expectation of performance. If you want them to run exactly as well on recovery as they did in the protected site you’ll need a fully dedicated set of resources that lines up 100% with what you have at the protected site.
2) During tests it will take snapshots, correct – the nature and size of those snapshots will depend on how long you intend to run the tests, how much data is changing during those tests, and the vendor and implementation of replication. Sorry I can’t give much more guidance here!
3) If you *can* leave the IPs the same it is better: Less chance for applications to break, and faster to recover without having to change all the addresses!
Hi, I just read texts on your blog and I became interested in the topic. I love your blog and I am thinking whether I could use your words in my paper? Might it be possible? If yes, please write to me.Thank you.
Can we use VMWare Replication to move VM’s permanently to a new Site (Cluster). We have a move to a new site coming up and the site will have all the hardware and infrastructure at the new location. We want to minimize the risk of moving our old equipment on move day by having the VM’s replicated to the new site. Then simply flip the switch. Is this a viable method?
R.
Yes indeed – Set up the replication of all the VMs you want to move, make sure they are replicated 100%, power them off and then do a recovery including replicating recent changes. You’ll need to choose which network to attach them to at power-on time.
Without SRM you’ll need to do this one VM at-a-time, and you’ll need to make sure they can still communicate on the network correctly when they power on, but it sounds like a perfect use case to me.
Find out more details by going to this web page which explains anti-malware in addition to the redirect virus by google. Topic :- Computer Safety
After looking into a few of the articles on your web site, I honestly like your way of writing a blog. I added it to my bookmark website list and will be checking back in the near future. Take a look at my web site too and tell me how you feel.