After a busy week in Vegas, a delayed flight home and less than two days to spend with the family, I’m off to VMworld again. Let’s talk about VeeamON before the next hectic week starts, tough.

On the conference, parties and amenities

As for VeeamON, Veeam’s first conference: I’m impressed. The general feel of the event is very good and everything was taken care of in brilliant fashion. As a blogger and speaker at the event, I enjoyed VIP perks at the event, but those weren’t what made the event special for me.

Instead, it was the very high production value of the content: the keynotes were ‘up there’, there were a lot of great sessions, the videos very professionally done and on a whole, every detail seemed well-thought out. I feel I really got to know Veeam as a company and the people behind it.

I liked the fact that we saw a lot of Ratmir on stage. Ratmir Timashev is one of those guys who’s not in the spotlight a lot, and the fact he was now is a prime indicator of the message they’re trying to convey. The late night talk show style of the keynotes were a nice touch that allowed the audience to get to know Ratmir a bit better, and perhaps most importantly, made Ratmir feel at home on stage. He was nervous, and that was totally o.k.. Better to have a slightly nervous and imperfect keynote than it being too polished.

I didn’t really like Beardy Joe and his friend Bob, though. I felt that this way of storytelling was a bit demeaning to the attendees, to be honest. I might have made a joke or two about those guys.

Screen Shot 2014-10-12 at 11.32.25

My break-out session

My coworker and me presented a community-led session on Wednesday morning. This hangover slot, 8:30 am after the Veeam Party (which was awesome, for lack of a better word) was our moment to shine. Despite the low number of attendees, we still had a blast doing our session on how to use Veeam in a Nutanix-based multi-tenant IaaS cloud. We shared a lot of details on how to deal with the new distributed architecture Nutanix brings to the table, discussed caveats when automating Veeam using vCloud Automation Center and doing self-service restores in a multi-tenant environment. Finally, we shared the nitty-gritty around using Veeam to back up meta data when using NSX. All in all, we had a good session with great interaction with the crowd. I will post a follow up with details how to watch the session online as soon as it’s been published.

From a technical perspective, there was a lot of new stuff being discussed. Highlight was of course Veeam Availability Suite v8, and it really got turned inside out during multiple sessions. But not only that, Veeam did a ‘one more thing’ trick and waited until the last day to publicly announce Veeam Endpoint Backup, which is a pretty big deal for me.

Veeam Endpoint Backup

So, with Endpoint Backup, Veeam’s finally stepped into the physical world. Even though it’s a small step as the product is specifically aimed at the IT Pro, I’m guessing that that’s not going to be the most popular use case. The real big news here is that this free tool is able to back up physical Windows servers. This enables customers with partly physical, partly virtual environments to stick with a single backup solution.

Even better, the tool will use existing Veeam technologies to do the backup, i.e. image-level backups (with volume and file granularity) and save it in the same format as a regular Veeam backup. Backup files can be read and restored using Veeam Backup & Replication free and paid editions, and all the Veeam Explorers work as normal.
Finally, the tool can also back up to an existing Veeam repository.

In other words, it’s all the good Veeam stuff you’re used to, but for physical servers. It does lack centralized management and other enterprise-grade features, but for a free tool, we shouldn’t complain, especially since I suspect this tool will, for now, be mostly used for those few remaining physical servers in an otherwise completely virtual world.

I’m sure this is only a very first and relatively small step for Veeam, testing out the waters and determining future direction of physical backup based on customer and partner feedback. I suspect centralized management and better integration into their portfolio are going to be next.

Veeam Availability Suite v8

Mantra for v8 and the re-named product suite is ‘people don’t need backup, they need availability’. Backup (and recovery) is just one small mechanical piece in the bigger ‘availability’ puzzle. They still focus on standard (Microsoft) applications running on a VMware or Microsoft hypervisor, which I think is a good thing; they keep doing what they do best and work to improve availability of those applications while being (mostly) agnostic about it.
Instead of going into the parallel processing, application mirroring or other application-level availability mechanism, they’re instead moving more and more into the lower layers of the infrastructure to improve recoverability. Key points here are adding more storage snapshot providers (NetApp, Data Domain) in addition to HP, integrating their data movers with physical repository appliances such as Exagrid and massively expanding off-site replication, adding failover (test) plans and into service-based Disaster Recovery with Cloud Connect.

Another cool enhancements is Backup I/O Control. This helps ensure performance on the primary storage platform when doing backups. Remember how backups could once be such a burden on production storage? I’m very curious how this’ll work in a distributed architecture, for example with Nutanix… I’m gonna have to try this out in the lab!

Finally, there are numerous small but very important feature enhancements around self-service restores, creating on-demand backups as a functional replacement for using hypervisor-based snapshots and more. They’ve also added a Snapshot Hunter to work around some of the issues with hypervisor-based snapshots.

All and all, v8 is going to be a huge release, and I’m hoping it’ll hit the shelves before the end of this year.

Concluding

Veeam’s first event seemed to be a successful one. There’s not really a lot to improve on, but there are a couple of points:

  1. More sessions, and more community sessions. For an event to be successful in upcoming years, the technical content needs to be expanded significantly. The quality of the sessions were already of pretty high quality, but in order to attract the an even bigger crowd next year, they need to diversify and expand the technical content.
  2. More community I was very glad to be able to deliver a community break-out session during the event, but sadly, the community speaking slot was at the morning after the Veeam Party, at 9.30 am. I would recommend Veeam to bring in more community speakers next year and give them better speaking slots

I’m very curious to see what Veeam will do next year with their VeeamON conference to convince this year’s attendees to show up again. Actually, Jim Millard posted a pretty conclusive post on this: VeeamON 2014: A post-event challenge, where he goes into the dilemma Veeam faces for next year’s conference.

Disclaimer: I was invited by Veeam to both present a break-out session and attend as a blogger. Veeam has paid for my travel, hotel stay and conference fee, but I am in no way obliged to write about the event.