Citrix screwed up recently with VDI-in-a-Box, creating a worldwide outage for every user of the product on January 1st. The proximate cause was a simple one, the Virtual Desktop Agent used by VDI-in-a-Box had a .dll which expired at midnight December 31st 2011 and the software wasn’t patched to address this. Like I said – embarrassing.
Now although it was a VDI product that had this problem wasn’t a VDI problem, it could have hit almost any product. I’ve seen this same type of failure take out many tens of thousands of PCs in one go and I’m sure it will happen again; people make mistakes. But once the problem was identified anyone running VDI-in-a-Box was able to implement a workaround and subsequently apply a permanent fix by touching a handful of servers in the data center rather than every single endpoint across the organization.
Quality control was the problem, rapid remediation was only possible through desktop virtualization.
What was the point of this post? If you wanted to point out a fix for VDI-In-A-Box it would have been better received than pointing out that the hand that fed you for so many years “screwed up”.
@Adam – If you still need to know how to fix this problem over 2 weeks later, you’ve got bigger issues than this.
The point if it needs to be hammered home is that this wasn’t a VDI problem, it shouldn’t be used to dismiss VDI but to show VDI’s benefits when it comes to recovering from a major failure.
@Simon – Please keep biting the hand that feeds/fed you. Some of us appreciate it.
Adam
You might not like how I described this incident but you can’t deny that it was a significant error on Citrix’s behalf. If Guise’s comments are correct (I don’t have any information to confirm this or not) and a fix was released to Citrix’s partners but not implemented in its own product then I think calling it a screw up understates the level of failure.
Yes, you are right I’ve done a lot of work with Citrix over the years, and the have fed me on occasion, sometimes outstandingly well. But I would like to think that on balance I have given as much as I have received, perhaps even more. Regardless the point of the post remains unchanged – this was not a problem with VDI and the fact that it occurred on a VDI system made it possible to recover from the incident far more quickly that it would have been possible to do so with any other desktop management technology.
Regards
Simon
Embarrassing is not the word, imagine starting the new year with absolutely all your hosted virtual desktops down.
Also this was not a specific VDI-in-a-box error, Kaviza got thrown under the bus on this one, it was actually a HDX licensing ‘glitch’ that affected anyone who is not Citrix but uses the HDX stack to deliver desktops.
We were not the only ones whose desktop estates were down for a day, the only people not affected were those running XenDesktop rigs and IBM Global Services (Desktone/HDX) who spotted the HDX license expiry code a few months back and had Citrix remove it.
So I am wondering if IBM Global drew their attention to this months ago, why on earth was this expiry code not taken out of the Kaviza and Desktone stack until it blew up on us ?