Parallels Cloud Server 6 : To upgrade or Not to upgrade?
Ex-colleague: “Parallels has released a new product. It’s superb!”
Me: “What’s it called?”
And this is how I first came to hear about PCS6 aka Parallels Cloud Server 6. This conversation occurred about a week before PCS6’s official launch on November 6, 2012. For the remainder of our chat, my ex-colleague went on to extol the benefits of the software. Truth be told, he really had me interested in this new direction that Parallels seemed to be headed in.
Parallels Cloud Server, for people who are new to the concept, combines the features of Parallels Server Bare Metal, Virtuozzo and Cloud technologies. So what you have is the ability to store and run Virtual Machines and Containers with the added advantage of distributed cloud storage. Simplifying it even further for emphasis, that’s hardware virtualization, OS virtualization and cloud storage all in one. I think that we have a real beast on our hands here.
How does this affect Web Hosting Providers?
Well for one, knowing Parallels the way I do, it appears to be only a matter of time before they stop upgrading their Virtuozzo and Server Bare Metal projects and ask all of us to move over to PCS. And then there is the “small” task of upgrading to PCS6 from our present setup, be it Server Bare Metal, Virtuozzo or OpenVZ. Thankfully, Parallels has provided detailed documentations for the same which can be easily accessed by browsing over to their website.
But, is it worth the trouble?
Well, it is pretty obvious that PCS6 arms Web Host Providers with a utility belt full of goodies that can greatly enhance and simplify business operations for any Web Host Provider. The main advantage cannot be stressed enough though. You can run both containers and hypervisors on the same physical server. Other key features include:
Distributed Cloud Storage
PCS employs replication to protect against temporary unavailability of containers and/or virtual machines by instantly migrating to an operational physical server. Another benefit is that it prevents data loss due to the presence of multiple copies.
Rebootless Kernel Update
Rather than having to reboot your container and/or virtual machine after a kernel update, these are briefly suspended during the update process. This can be attributed to two main differences during the kernel load operations:
- Hardware does not need to be reinitialized.
- Container’s and/or virtual machine’s private data is preserved.
Container In A File(containers only)
Using ploop aka Parallels Loopback Device, all of the container’s data can be stored in a single image. This avoids conflicts between multiple container operations.
Robust Deduplication(containers only)
Files that are frequently accessed by containers are determined. And from this list, files that are commonly used by all containers are placed in cache. This improves performance by liberating IO cycles.
Sum Of All Things
When it comes down to whether or not you should go for it, it depends entirely on your environment. If you are new to this industry and if you are searching for a platform to propel your business forward; in my opinion, PCS is the right way to go. Owing of it’s flexibility, the options at your disposal are numerous.
However, if you are already using either Parallels Server Bare Metal, Virtuozzo or OpenVZ; I guess the title of this blog is not apt for your predicament. The question should not be “IF?”, the real question should be “WHEN?”