Category: News

Why VMware VVols are Simpler, Smarter and Faster then traditional storage

Recently at VMworld in Vegas, Pete Flecha from VMware did a presentation on VVols at the HPE booth which highlights the benefits of VVols being simpler, smarter and faster than traditional storage with vSphere. The link to the presentation is below but I thought I would also summarize why VVols are simpler, smarter and faster.

VVols are Simpler

The reason VVols are simpler is because of storage policy based management which automates the provisioning and reclamation of storage for VMs. VVols completely eliminates LUN management on the storage array as vSphere can natively write VMs to a storage array, VVols are provisioned as needed when VMs are created and automatically reclaimed when VMs are deleted or moved.

VVols are Smarter

The reason VVols are smarter is again because of storage policy based management and also by the dynamic real time nature of storage operations between vSphere and the storage array. With VVols a storage array stays as efficient and thin as possible as space is always reclaimed immediately for VMs and snapshots and even when data is deleted within the guest OS it can be reclaimed on the storage array. In addition the array manages all snapshots and clones and array capabilities can be assigned at the VM level with SPBM.

VVols are Faster

The reason VVols are faster is not what you might think, data is still written from vSphere to a storage array using the same storage queues and paths, there is no performance difference between VVols & VMFS for normal read/write operations. Where VVols is much faster than VMFS is when it comes to snapshots, because all vSphere snapshots are array snapshots they are much more efficient and they take the burden off the host, In addition because there is no need to merge any data that has changed while a snapshot is active, deleting snapshots is always an instant process which can have a very positive impact on backups.

 

Go watch the whole video to learn more about why VVols are so great , it’s only about 15 minutes long.


Share This:

Nominations for the 2018 Top vBlog voting categories are now open!

Just like prior years in addition to the traditional voting I’m opening it up to allow voting in specific categories as well to help distinguish certain types of blogs. To do this I have created a form to allow you to nominate your blog or website for one of the categories that I have defined if your blog fits one or more of the categories, if it does not then don’t nominate it as all blogs on the vLaunchPad will automatically be included in the general top blog voting.

This form is not the general voting poll for the top VMware/virtualization blogs, this form is only to nominate your blog for certain categories if it fits. Once the nominations are collected I will open the polls for voting for the top blogs where voters will be able to rate their top 10 blogs and also vote in each of the categories.

You should only nominate your own blog/website, these nominations will be used to populate the category choices when voting opens. If your blog doesn’t fit one of these categories then do not nominate it, all blogs on the vLaunchPad will automatically be included in the general top blog voting. If your blog is not currently listed on the vLaunchPad use this form to let me know.

The criteria for the categories to see if your blog qualifies for them are as follows:

  • Favorite New Blog – Blog must have been started in 2017
  • Favorite Storage Blog – Must have greater than 50% posts as storage related
  • Favorite Independent Blogger – Must not work for VMware or a Technology Partner (i.e. EMC, Dell, IBM, Unitrends, etc.), basically this means customers, resellers and integrators only.
  • Favorite Scripting/Automation Blog – Must have greater than 50% posts as scripting/automation related
  • Favorite Non-English Blog – Must be primarily published in a non-English language
  • Favorite Podcast – Easy one, you must have a podcast
  • Favorite Female Blogger – Special recognition for women in tech, must be a female blogger
  • Favorite News/Information Site – Any site that primarily publishes news and information and is not really a blog

Again, only nominate your own blog/website!

Share This:

Everything you need to know about UNMAP in vSphere

As I watch the VMworld session recordings which are all publicly available I’ve been doing a write-up and summary of those sessions. My write-up is only a small summary of those sessions which are usually packed full of great information so I encourage you to go watch the full session recording. Today’s topic is UNMAP, a feature I have been very involved with since it’s initial release and a feature that has greatly evolved and changed across vSphere releases. John & Jason from VMware’s tech marketing do a great job talking about the history of UNMAP, showing examples of UNMAP in action and making recommendations for getting the most out of UNMAP. Below is my summary of the session and some highlights from it:

Session:  Better Storage Utilization with Space Reclamation/UNMAP (HCI3331BU) (View session recording)

Speakers:  Jason Massae, Technical Marketing Architect, Core Storage, vSAN, VMware – John Nicholson, Senior Technical Marketing Architect, VMware

  • The session opens with a description of space reclamation and the different levels it can be performed at and explains why it is an import feature. Next they go into the history of UNMAP which was a bit troubled at first, it debuted as part of vSphere 5.0 as an automatic process but issues with some vendors being able to support it properly quickly surfaced and in 5.0 Update 1 it was disabled by default. From that point on it became a manual (CLI) process until it finally came back in vSphere 6.5 as an automatic process with some modifications to force the process to work at fixed rate levels. In 6.7 they further refined UNMAP with a configurable rate to provide more flexibility and in 6.7 U1 vSAN it became it a truly automatic process again (for vSAN only).

  • Next they go into more detail on how the process worked across different vSphere versions. I went into a lot of detail on this in this post I did on UNMAP in vSphere 6.5, in 6.7 the only changed was going from fixed preset limits to more flexible configurable limits. They didn’t mention it but take note that UNMAP with VVols has been automatic since vSphere 6.0, a host doesn’t need to tell an array which blocks to reclaim when a VM is deleted the array is already aware of it and it’s on the array to do the reclamation on it’s own. Because of this flexibility an array can hold off on reclaiming space and potentially allow a user to undelete a VM if needed (i.e. recycle bin).

  • Next they cover in detail what was actually un-mapped in different versions of vSphere, 6.0 was pretty limited to thin disks only, in vSphere 6.5 it became less restrictive and now with vSphere 6.7 it pretty much works with everything. One thing to note again with VVols is that un-mapping from within the guest OS is supported and also snapshots are automatically reclaimed as well when they are deleted as all snapshots with VVols are array based snapshots.

  • As far as what types of Datastores are supported with UNMAP, what it really comes down to is how the UNMAP is handled. With VMFS vSphere tells the storage array what blocks to UNMAP as the array had no visibility inside the VMFS volume and doesn’t know where VM data resides. With NFS which isn’t a block based file system the array is already aware of which disk blocks a VM is on as a VM is written as a file, once the VM is deleted it knows which blocks to reclaim when deleting that file. The same holds true for VVols, the array knows where the VM is written and just deletes and reclaims the VVols associated with a VM. It’s VM level visibility great?
  • From there they went on describing the mechanics of how UNMAP works and some best practices for using it effectively. They also showed how to monitor the performance impact of UNMAP along with some demos. I just touched on a small part of this session so I encourage you to go watch the session replay to learn a lot more about UNMAP.
Share This:

VMware VVols from a customer perspective

What do donuts and nuclear power plants have in common beyond the obvious answer? (hint: The Simpsons) They both are companies that are running VMware VVols! At VMworld this year instead of VMware putting on a VVols partner panel they chose to do a customer panel session instead. This was a great move as it lets you hear direct from customers actually using VVols on their experiences with VVols and the benefits that they found with VVols. Below is my summary of the session and some highlights from it:

Session:  Leveraging Virtual Volumes (VVol) to Simplify Storage Management (HCI2550PU)  (View session recording)

Speakers: Bryan Young, Group Product Manager, VMware – JJ Seely, Sr. Server Administrator, NuScale Power, LLC – Michael Bailess, Infrastructure Technology Manager, Krispy Kreme Doughnut Corporation

  • The panel was moderated by Bryan Young from VMware who is the product manager for VVols and consisted of 2 customers, J.J. Seely from NuScale Power and Michael Bailess from Krispy Kreme donuts.

  • Bryan first gave an update on where VMware is at with VVols, he showed off the timeline of VVols from it’s initial release as part of vSphere 6.0 in March 2015 to the vSphere 6.5 update in November 2016 which brought the long awaited support for replication and finally to the most recent release, vSphere 6.7 in April 2018 that added a few more VVols enhancements

  • Next he made the point that despite you not hearing too much from VMware anymore about VVols, it is a very active program and growing significantly. Over the last few months they have seen a 2x increase in the number of deployments of VVols and a 3x increase in the capacity of VVols deployments, much of that happening in the last 6 months. In case you are wondering they can track this with vCenter Analytics Cloud data (phone home). The partner commitment around VVols remains strong with many partners working to finishing their VVols 2.0 (replication) solutions. On VMware’s side they are growing their engineering team to support VVols and have a solid roadmap going forward for enhancing VVols.

  • VMware’s partner ecosystem for VVols has been growing and every major storage vendor at least supports VVols 1.0 today. As each partner is at different stages of completing their VVols solutions it’s best to reach out to them and find out their roadmaps for supporting VVols.

  • Bryan reviewed what VMware delivered in vSphere 6.7 around VVols, the big one being support for  SCSI-3 Persistent Reservations. Why is that one big? Because there is no more need for RDMs as VVols can support this natively. Many customers are still using RDMs to support Microsoft Windows Failover Cluster Server which is really the only use case for RDMs. Nobody really likes managing RDMs so it’s great that VVols can do this natively now. I actually know a few customers that have mass migrated their RDMs to VVols.

  • Bryan then highlighted an important aspect of VVols that many people don’t get. VVols & vSAN are not an either/or decision, they work perfectly together as they are both based on the same Storage Policy Based Management engine native to vSphere. There is seamless interoperability between VVols & vSAN and VMs can move freely from one to the other.

  • From that point each customer did a short presentation and there was a lot of Q&A from the attendees. Go listen to the recording and find out what they had to say. You can also listen to a separate breakout session where Michael Bailess from Krispy Kreme went into his experience with VVols in more detail.
Share This:

SRM deployed as an appliance

One of the VMware sessions on SRM at VMworld which you can view online right now featured a tech preview of SRM deployed as a virtual appliance instead of being a Windows based application. Apparently this new SRM appliance model already exists and is used in VMware’s SRM on AWS offering. I’m guessing that VMware’s motivation for moving SRM off Windows was initially to support AWS which required SRM to be in a cloud friendly format. They must off decided that since they did that they might as well bring it to the on-prem version of SRM as well.

Of course this will have some big benefits for SRM customers such as no more Windows OS to license at the primary and secondary sites. It also greatly simplifies deploying and upgrading SRM and is also less Windows servers to maintain and patch. It should also make it easier for vendors to package and deploy their SRAs for SRM. Although at some point once they support VVols on SRM their will be no more needs for SRAs.

I just watched the session and thought I would provide a summary and some highlights from it.

Session:  Site Recovery Manager 8.1: What’s New? (HCI2474BU) (View session recording)

Speakers:  GS Khalsa, Senior Technical Marketing Architect, VMware – Velina Krasteva, Product Manager SRM, vSphere Replication, VMware

  • First as with any tech preview VMware provides, they are not committing to any dates or releases, so when it comes it comes you’ll just have to be patient
  • The SRM application will run on VMware’s Photon OS and the SRA’s will be delivered and packaged by vendors inside a Docker container

  • The SRM appliance will be installed from a standard OVF template, you fill in the basic configuration for the appliance and it then gets deployed

  • Once installed you can access the configuration screen for the SRM appliance and configure things like the PSC, vCenter credentials, configure SSH access, time, syslog and most importantly install SRAs

  • Now you can install an SRA by browsing and selecting the Docker image that each vendor will provide

  • Once the SRA is installed you can see the details of the SRA

  • And finally by connecting to the appliance via SSH you can view the SRA installed as a Docker image

Share This:

My thoughts and observations on VMworld 2018

It’s that time of year again when 20,000+ virtualization geeks get together in flashy Las Vegas to pay homage to VMware at VMworld. This year made #11 for me, it’s hard to believe I’ve been going to the same conference every year for over a decade. One thing that never changes is that there is never a dull moment and it always goes by so fast despite being there for 5 days. As always it’s great to see old friends and also meet new ones, even with 20,000 people there I always seem to constantly run into people I know all the time there. Now for my thoughts and observations on VMworld 2018:

vExperts at VMworld

I made my special magnetic vExperts badges again this year to hand out, both the vExpert 2018 logo and the 10 year vExpert logo. I enjoy making them and giving back to the community. This year there were not many vendors doing vExpert giveaways, Datrium was the only one that I saw doing it, thanks Datrium for the cool hoodie. I think that with the large number of vExperts it gets pretty costly for vendors to do this so most don’t anymore. I’d like to see vendors maybe target top bloggers instead as it’s a much smaller group. Thanks to Runecast that did just that this year with a wicked cool shirt (see below). As always thanks to VMware, specifically Eric Nielsen and Corey Romero for taking care of vExperts at VMworld with a swag bag and a special party where Pat always makes an appearance.

This year the party was again at the Pinball Museum which is just off the strip in a fairly nondescript run-down building. The inside is completely filled with pinball machines and classic electronic games, out back they had some tables and a caterer cooking some BBQ. Pat made his appearance and was mobbed by people, let me tell you that guy is an absolute saint, so easy going and patient and makes time for everyone, talking to him you feel like you’re talking to a fellow virtualization geek and not a top executive. I got my annual pic with him, we had a 10 year vExpert informal pic also and he rolled up his sleeve and showed us his tattoo. Let’s talk about that for a minute.

About that whole vInk thing…

Pat showed off his newly obtained tattoo in the first day keynote. I started to write about it here but it got too long so I did a separate post on it.

Location

Once again VMworld was in Vegas at the Mandalay Bay, alas this is the last year there as Moscone Center is ready to welcome VMworld back after years of renovations. It will be a nice change being back in San Francisco after several years of Vegas but I wish they would go back and forth as Vegas is built for conferences and San Francisco is not. They can do all they want to make Moscone better for large conferences but that does nothing to fix the crappy hotel situation there.

This year I stayed at Excalibur after taking a pass on our room block at the MGM Signature. It’s all about location for me and Excalibur is physically attached to Mandalay Bay as well as has a tram directly to there. Mandalay Bay is a good location for the conference and does a great job accommodating it. I really have no issues with Vegas, it’s up all night, airport is real close, cheap and plentiful hotel rooms and it’s more centrally located then the Bay area.

How many people attended VMworld?

About the same as last year, VMware reported 20,000+ in the keynote which indicates it was over 20,000 and under 21,000. At least it is consistent and not trending downward but it is down from it’s peak of 23,000+ 2 years ago. Should be interesting to see what it is next year when it moves back to San Francisco. The attendance of people at your conference can be an indicator of the health and popularity of your business. VMware has never stated what that mix of total attendees is comprised of but I can guess that it includes VMware employees, partners & booth staff, press and analysts and customers. How much of each is anyone’s guess, I’d say at least half of that number is made up of customers.

What was announced at VMworld?

This year the event didn’t coincide with a major release of vSphere, there is 6.7 Update 1 coming and it mostly has vSAN enhancements in it. There are also some updates to the companion vRealize products but most of the announcements were cloud centric. One thing of note with 6.7 U1 is that the HTML5 UI is now fully functional instead of being 95% functional. Yes you can now ditch your Flex client for good as long as you upgrade to 6.7 U1. There it also a new vCenter Converge Tool which makes migrating from an external PSC to an embedded PSC much easier. You can read all about what’s new in vSphere 6.7 U1 here. Also you can watch the VMworld breakout session on what’s new here. Note that 6.7 U1 was just announced and not yet available.

VMware also announced a new vSphere edition! Just what we needed, make licensing more complicated than it is already. The vSphere Platinum edition a step up from Enterprise Plus and basically just adds security features (AppDefense). No word yet on cost, it is still licensed per CPU but you can bet it will cost more than an Enterprise Plus license. You can read more about what you get with Platinum Edition here.

On the cloud side VMware announced expansion of it’s offering on AWS to Asia-Pacific. This makes them globally complete except for Antarctica, they expanded to EMEA back in March. They also had some announcements around NSX including more integration with AWS. One of the more interesting announcements was around a new Relational Database as a Service offering on AWS. This is basically a similar offering to what Amazon already provides just running on the VMware stack. This includes support for most major databases including Oracle, MS SQL, MySQL and PostgreSQL. According to VMware this will be generally available soon and you can register for a preview of it here.

They also announced Project Dimension which is essentially an on-prem to edge hybrid cloud offering. It includes VMware Cloud Foundation coupled with NSX SD-WAN by VeloCloud to enable connectivity with other remote locations coupled with a unified hybrid cloud control plane. Doesn’t really sound like anything new, more just some repackaging of VMware products. You can read more about it in this blog post.

Another announcement was on VMware Cloud Automation Services, again this is just re-branding and re-packaging of VMware products. The cloud automation services consist of VMware Cloud Assembly, VMware Service Broker and VMware Code Stream. Continuing on the cloud theme that also announced they were acquiring CloudHealth and make it a VMware branded offering while integrating it with other VMware products like WaveFront.

ESXi on ARM was another interesting announcement, this would allow ESXi to run on that lightweight CPU platform allowing them to extend  ESXi to the millions of IoT and edge devices out there. They really didn’t show or talk about this too much so I assume this is just a sneak peek and will eventually be supported at some point. As ARM devices are typically low cost the VMware license would have to be either free or very low cost as well.

How were the General Sessions?

I didn’t attend either and instead watched them from the live stream. Last year I found the general sessions to be good and interesting, I like the whole fictitious Elastic Sky Pizza narrative they did which was very creative and fun. This year I found the general sessions mostly bland and rather boring. It seemed like VMware didn’t have much to talk about and mostly re-hashed their current offerings some with new names. The keynotes were mostly cloud centric as well which is were much of VMware’s focus is these days.

The Day 1 General Session was mostly cloudly, opened with Pat celebrating VMware’s 20 year anniversary and how the community was such a big part of that. He then showed off his newly obtained VMware tattoo, then talked about the VMware vision and where we are today compared to in the past. He talked about how AI & Machine Learning will play a big part of VMware’s future product offerings and then switched to cloud mode more specifically about their AWS offerings. He brought out Andy Jassy, AWS CEO again this year and talked briefly about their partnership. Then Ray O’Farrell came out and dove deeper into the clouds talking about migrations using HCX and Project Dimension and VMware PKS. Next they did a demo of the vGPU vMotion feature, they also briefly showed off the ESXi on ARM upcoming functionality. Finally they showed off Workspace ONE, NSX & AppDefense. No sign of Michael Dell in the keynote this year.

The Day 2 General Session was all Sanjay talking for about 10 minutes and then interviewing people, first several customers, then a few partners and then finally Malala Yousafzai who is a Pakistani activist. I found this highly boring, I only skimmed through the replay of it which does not include the interview with Malala.

How were the Breakout Sessions?

I only attended 2 and presented in 1, there were a few more VVols sessions then last year. I went to the Patrick & Pete VVols deep dive session which was largely a repeat of prior years but was still very good, it had pretty decent attendance, maybe around 300+. I also went to the VVols customer panel session hosted by Bryan Young from VMware. He had 2 VVols customers there to present and tell about their experience with VVols, he also had some good Q&A from the audience, there was maybe 100+ at that session. The session I presented at was an HPE sponsor session which was largely a marketing session but actually had a decent turnout (100). Another interesting session which I would recommend is the What’s New in SRM 8.1 session which features a tech preview of the new containerized SRM architecture coming soon.

This year the sessions have all been posted for any one to access already. VMware hasn’t posted anything to YouTube but you can watch them all at the VMworld On-demand Video Library. Some of the more popular sessions were recorded with video, were the other sessions are just audio and slides. I still need to go through and see what I missed and watch a lot of the replays, as always you’ll find lots of great content in there.

What was going on in the Solutions Exchange?

Really it was almost an exact replay of last year, same layout and same vendors with their flashy booths had them back again. Rubrik had the basketball court again and an NBA player on-site, IBM had a skateboard theme going and had Tony Hawk in booth and there were a lot of other vendors that had creative and flashy booths to draw attendees in. I thought there was good traffic in the Solutions Exchange, really about the same as last year. The Solutions Exchange tends to be the area were I spend most of my time at, it’s always fun to walk around, meet people and check out all the products on display.

How about the parties?

There seemed to be distinctively less parties this year, I can understand why as vendors are less willing to dump money into parties as there is very little ROI on them. I went to VMunderground again which was decent, the Zerto party was pretty good also although I enjoyed the Journey tribute band more last year then the Pearl Jam tribute band this year. There was also the usual big parties, Cohesity had Snoop Dogg, Rubrik/AWS had RUN DMC and Veeam had their big night club party. I skipped most of those as I hate the night scene and prefer more intimate parties.

Final Thoughts

After 11 years attending VMworld I still enjoy it, VMworld to me it’s about the parties, it’s not about the keynotes nor the sessions, it’s all about the interaction with other people attending the event. I thought the venue was great, the whole event was well executed, the VMworld app seemed more useful this year and overall I thought VMware did another excellent job pulling it off. Despite being there for 4 days I always leave with a sense of regret and wishing I could have done more or had more time. It almost felt like breaking up with an old friend when it was time to depart. I look forward to going to SF next year for #12 and now I’ll spend the next few weeks reviewing all the session content I didn’t have a chance to see onsite.

And now some pics:

Thanks Runecast for the badass shirt!

The iconic VMworld sign at the entrance to the VMvillage

The annual Pat & I selfie

VVols Deep Dive session

Pearl Jam tribute band at Zerto party

10x vExpert with Pat group photo at vExpert party

Rubrik booth

vmVillage area

VVols customer panelPete Flecha preaching the VVols gospelAnnual Calvin, Angelo and I selfie

Share This:

About that whole vInk thing…

Pat showed off his newly obtained tattoo in the first day keynote, he was wearing a sport coat to begin and they went to play a video to celebrate VMware’s 20th anniversary and once they came back first thing I noticed was he looked different (no sport coat). He then rolled up his sleeve and displayed a pretty sizeable VMware tattoo on his arm. He said he got it to take his commitment to VMware to the next level and came into town a few days early and went to Bad Ass Tattoo to get it. He also said sometime what happens in Vegas (tattoo) doesn’t stay here.

Now everyone’s immediate reaction was, holy cow is that real? He did say “doesn’t stay in Vegas” which indicated it was permanent but that was quite a big tattoo in a very visible spot and to have that for life is more than just taking commitment to the next level, that’s going right to the top level. Later that day he also tweeted out a pic of him actually getting the tattoo.

I’m still undecided myself, it could be a marketing stunt, but who knows. I saw Pat in person the next day and he rolled up his sleeve and it wasn’t faded at all so it very well could be real.

Being the overly curious type though and always wanting to solve a mystery I investigated and here’s what I found.

  • There is no tattoo shop called “Bad Ass Tattoo” in Vegas, now he might have just been generalizing instead of wanting to name the place specifically.
  • The picture of him getting the tattoo wasn’t at a tattoo shop, it really looked like a hotel room to me. I figured he would be staying at the Delano or Four Seasons and sure enough I matched the room decor in the picture of him getting the tattoo with a picture of the dining room in the Presidential Suite at the Four Seasons hotel from TripAdvisor (see comparison below). Now if he was to get a tattoo, I could see him wanting to do it in a private location like a hotel room so doesn’t necessarily mean it was a staged picture. However the picture showed the tattoo finished and not as a work in progress so it could of very well been staged.

  • If it was indeed a temporary tattoo, it was no way applied with a tattoo gun, all tattoo’s done by a tattoo gun’s needles are permanent. Temporary tattoo’s are all applied by printing out an image on transfer paper and using water to apply it, they generally last 1-5 days. It could also have been airbrushed which last about a week or even done using a sharpie which typically fade after a few months.
  • If he had it done a few days beforehand the area would typically be red or flaky, it looked completely healed which typically takes at least 2 weeks. Normally as the skin heals it may be red and oozy for the first week and eventually flaky after that.

So real or not? I can’t say for sure, it’s entirely possible he got the tattoo before coming to Vegas to give it enough time to heal and then staged the part about getting it there. It’s also entirely possible that it was done with a Sharpie that would of lasted the whole show without fading.

I guess we’ll have to wait until VMworld next year and see if it’s still there to know for sure. One thing I do no doubt one bit is Pat’s commitment to VMware, I don’t think anyone could question that. VMware is lucky to have such a great leader and I could definitely see how he might want to replicate the mark he’s made on VMware onto himself.

You can read more about Pat Gelsinger in this blog post I did last year.

Share This:

VMware finally announces SRM support for VVols!

VMware’s new storage architecture, Virtual Volumes (VVols), has been part of vSphere for over 3 years starting with the vSphere 6.0 release. However the initial release did not support array based replication which VMware eventually provided in vSphere 6.5. That support came with a caveat though, while replication was supported via Storage Policy Based Management (SPBM), orchestration of replication operations was a manual and painful process and not supported with SRM. To do a test failover, failover or failback you had to use PowerCLI and write your own scripts to orchestrate those operations, not exactly something you want to be dealing with in a time of crisis.

VMware SRM was designed to make BC/DR a simple and easy process and allow automated orchestration of replication by simply pushing a button. SRM essentially takes over the ownership and control of array-based replication using a Storage Replication Adapter (SRA) provided by each vendor specific to their storage arrays. Therefore when you click a button inside of vSphere, SRM has full control of array replication, bring up VMs at the recovery site and eventually failback to the primary site when needed by reversing replication. Having no SRM support for VVols is a show stopper for most customers that don’t want to deal with complex and manual scripts to perform BC/DR operations. Note you may have heard that SRM does support VVols today but that is ONLY if you are using vSphere Replication.

Let’s first examine how we got here and why it’s taken so long for SRM to support VVols. Most of VMware’s products are in silos, meaning they are run by different product teams, as a result product roadmaps and interoperability are not always in sync across products. Every product team has their own priorities and it may or may not include immediate support for other VMware products. VVols development has it’s own product team and they are mostly solely focused on developing VVols. Support for VVols with SRM is completely outside of the VVols engineering team and solely within the SRM engineering team. It’s not up to the VVols product manager to decide on when SRM will support VVols, it’s entirely up to the SRM product manager.

For the last year or so we’ve pleaded to the VVols team for SRM support and we were essentially told sorry, bring us customers that want it and we’ll try to push them to prioritize it. This of course frustrated everyone as support for replication without support for SRM wasn’t a complete solution. VMware’s dead silence on the issue also didn’t sit well with customers and left many customers not wanting to use VVols. Several months ago we had a face to face meeting at VMware with Lee Caswell and the product managers for VVols & SRM and again pleaded for VMware to take action. At that time they did finally commit to supporting VVols on the SRM roadmap but we also requested that they communicate that to customers sooner rather than later.

Well VMware finally broke their silence and announced SRM for VVols right before VMworld in a blog post. I’m betting the timing on this was deliberate as last year VMware got really beat up over this at VMworld and probably didn’t want a repeat of that. Note the blog post is very vague on purpose and this is just an announcement and don’t expect support for it right away. The SRM team is working on something else big right now which if you were at VMworld you could of seen a tech preview of it. The announcement basically acknowledges that VMware clearly sees SRM support for VVols as a key priority finally so customers and partners can put their pitchforks away.

Note with this announcement VMware is not just telling customers that support is coming so they can plan for it and start migrating to VVols, it’s also putting partners on notice to inspire them to prioritize finishing their support for VVols replication. To this day, almost 2 years after VVols replication was supported in vSphere 6.5 their are still only 2 partners (HPE/Pure) that support it. I think some partners may not have prioritized VVols replication support as they saw their was no SRM support for it and figured why bother.

So how will this marriage of VVols & SRM look? For one thing their will be no more SRA’s to install and maintain into SRM. Control of replication will be handled natively through the VASA Provider without any external components needed. This will be a welcome change and greatly simplify using SRM with external storage arrays. How this will impact certification (HCL) is TBD, presumably it will be handled through the VASA certification process. Beyond that we’ll have to wait and see, I know this is a nowhere near a minor change and represents a big engineering effort for the SRM team. I’m fortunate to be working closely with the SRM product manager and engineering team on this and am excited to see one of the final barriers for customers wanting to use VVols disappear.

Share This: