Skip to main content

Hello everyone! Got a couple of really cool announcements in the world of Google Cloud that I’d love to share with you.

Performing Mass VM Migrations with Google Cloud and Velostrata

One of the biggest challenges in moving to the cloud is migrating all of your existing servers without extensive downtime. But what if you could keep your servers running in production, during the migration process? Sounds like magic, doesn’t it?
Back in May, Google acquired a company called Velostrata, which specializes in enterprise cloud migrations. Since then, Google has detailed how they’ve used Velostrata’s capabilities to migrate huge numbers of servers to Google Cloud while running in production. It sounds like magic, but the solution is quite clever.

Unlike the typical approach of moving a copy of everything, the Velostrata approach is to instead separate storage from compute and migrate data transparently in the background while your cloud server is running.
Let’s take a closer look at how this migration process works:
First: An extensive discovery process takes place, which gathers information about the source servers to be migrated, and maps it to equivalent Google Cloud resources.

Second: Once discovery is complete, new Google Cloud servers are spun up and the actual data migration takes plac,e while both the source and destination servers are running. This only requires a short 10-minute downtime that is easy to schedule. Afterward, data is migrated while it’s also being served in production, and the migration process is mostly transparent to your cloud-based applications.

Finally: Postmigration observes the migrated VMs for a few days, just to make sure everything works correctly, then transitions to regular operations. At that point, you can shut down your original on-premises servers.
Overall, it is a fascinating process that takes the pain out of large and complex server migrations, and makes the standard lift and shift process a lot easier.

Building a Hybrid Render Farm on Google Cloud

Working with visual effects and animation in the entertainment industry takes a lot of computing power to make magic happen. Building, configuring, and maintaining the render farms to make this possible is quite difficult and expensive. In fact, many entertainment companies want to get out of the hardware business and would rather just focus on their work. Fortunately, Google Cloud makes that possible.
Last week, Google published an extensive guide on how to create a hybrid render farm that takes your on-premises files and uses Google Cloud resources for the heavy duty work of large rendering jobs. A variety of solutions are possible, depending on budget, location, and Internet speeds. Using the power of scalable cloud resources, as well as the ability to create single VM’s with up to 160 CPU cores and almost 4 TB of RAM, even the biggest rendering jobs can be completed quickly while saving costs in comparison to a permanent, on-site machine. We will link the full guide in the video description, so be sure to check it out.
That’s going to wrap up this edition of Google Cloud Weekly. Thank you for watching, be sure to subscribe to the Linux Academy YouTube channel for the newest Google Cloud updates, and I’ll see you next week. 

Check out the previous episode of Google Cloud Weekly (8/17/2018), and here’s the next episode (8/31/2018)



Leave a Reply

Your email address will not be published. Required fields are marked *