The Darwin Awards salute the improvement of the human genome by honoring those who accidentally remove themselves from it…

If there were a corporate Darwin Award given, Symantec would have to be a leading contender in 2012.

Someone at Symantec recently wrote a blog post calling out that they found a defect in the “vStorage VMware 5.1 API”, to wit:

Symantec and VMware are unwaveringly dedicated to quality in our jointly developed solutions. During our testing with the VMware vStorage 5.1 API Symantec encountered issues* that introduce risk performing consistent backups and ensuring reliable restores. We raised these concerns with our friends at VMware, who documented them in their knowledge base and in the VMware vSphere 5.1 release notes. VMware expects a vStorage API update, and at that time we expect to retest and once again support joint solutions.

During these weeks of close examination VMware also confirmed that any 3rd party vendor depending on the current API may be impacted by these issues. We can only conclude that compatibility and support claims from other vendors are either poorly informed or possibly represent a more cavalier “ship-and-fix” data protection approach than Symantec is willing to accept. […]

Another vendor took offense. My favorite part of their blog post was their calling Symantec on the carpet for its marketing campaign.

Symantec went on a campaign to scare people from using [our product] and more importantly to scare them from upgrading to vSphere 5.1.

So what’s really going on here?

First and foremost, let me note that I absolutely hate writing about this – and am doing so only because of the questions that have been raised by our partners, customers, and buyers. Symantec has had a truly terrible 2012 – between the BE2012 (Backup Exec 2012) debacle and the slowing sales and the firing of their CEO – well, it’s been the year from hell for those guys. Early in the year I was tough on Symantec – I stopped because even though they are a market leader in the backup space, they were a million times more effective than I was in pointing out Symantec’s weaknesses. Or to put it even more bluntly – if someone is continually shooting themselves in the foot you don’t really need to help them shoot faster.

Symantec shot themselves AGAIN in the foot here. They took the fact that they are late delivering support for VMware vSphere 5.1 and used a defect in the latest VMware Virtual Disk Development Kit (VDDK) to justify why they were late. And worse, they then mounted a marketing campagin explaining that all of us who said we supported VMware vSphere 5.1 were either stupid or didn’t care about quality.

To a virtualization-only vendor, this is life or death stuff. Virtualization-only vendors have only one way to backup – through the HOS (Host Operating System) level. There are strengths to HOS-level backup, and there are weaknesses as well (read Dogma, Faith, and Fact: Do You Have to Pick a Virtual Backup Religion for much, much more on this), but at the end of the day Symantec was saying that the ONLY way a virtualization vendor has to do backup didn’t work. No wonder they were spitting mad. Symantec would have been much smarter to point out that much like the man who has a hammer and therefore everything looks like a nail, virtualization vendors’ HOS-only backup approach may not be the most applicable to modern agile IT.

Unitrends understood the VDDK issue and took steps to make sure that we didn’t have the issue. Nothing made by man of which I’m aware is perfect – and VMware isn’t perfect (just as Microsoft, Red Hat, IBM, and others aren’t perfect.) Experienced software developers understand that and are capable of creating systems that work around problems.

At the end of the day, this is a tempest in a teapot. I think it says more about the marketing of most backup vendors than anything else.

Of course, that’s just my opinion. Comments, questions, and even criticism – more than welcome!


  1. <<“HOS” (how you call it), or much more industry-standard “host-based” backup is how VMware was backed up before VADP was introduced.<<

    Actually, before VADP was introduced you had VCB. But let's try to get beyond that.

    Here's the fundamental issue. If you're using VADP and you can *ONLY* use VADP, you don't have a whole lot of choice in how you backup VMware – whether on DAS, NAS, or SAN. If you're able to use VADP but also you can work directly within the virtual machine, you have a choice. Make sense?

  2. Folks, just to set the record straight: While there are plenty of editorial positions taken in this article about Symantec that I agree with, the idea that Sym was late with 5.1 support is 100% FALSE. The 5.1 VDDK was a mess and VMware promised to release a fixed VDDK quickly. They didn’t.

    You can argue that Sym should have just ripped out the new functionality from the update so that the buggy parts of the VDDK didn’t need to be leveraged (which is what they wound up doing anyway) but to argue that they used this as an excuse because they were behind is completely untrue.

    In terms of their strategy on this, that can be argued all day and I agree with much of what you’re saying. But it wasn’t a matter of them clutching at straws because they fell behind, the decision was made out of a genuine concern for customer data.

    1. Thanks tremendously for the note and for your insight here.

      Let me ask you a question: when you say

      “Folks, just to set the record straight: While there are plenty of editorial positions taken in this article about Symantec that I agree with, the idea that Sym was late with 5.1 support is 100% FALSE. The 5.1 VDDK was a mess and VMware promised to release a fixed VDDK quickly. They didn’t.”

      It would seem to me that what you’re positing is that Symantec was intentionally late with 5.1 support because it in their view VMware released defective software – and that what’s important to note here is that the delay was intentional rather than inadvertent. Is that a correct restatement of your argument?

Comments are closed.