For years our customers have told us that they understand the benefits of VDI and secure application and data delivery, but they need us to deploy and manage to make as easy as possible. That is why we have introduced the Citrix Cloud and its services. reduce complexity, and make it as easy as possible to implement Citrix technologies
The XenApp and XenDesktop service is a good example. It is a Citrix cloud offering, where we host the infrastructure components. If the XenApp and XenDesktop service implementation, we wanted to develop a cost-effective way to manage Delivery Controller for our customers. Delivery Controllers are an important part of any page, and we wanted consistency across all delivery controller instances to ensure running in Citrix Cloud. To do this, we keep all virtual machines hosting the Delivery Controller requires up to date with the latest security patches.
We knew that as more and more cloud-hosted XenApp and XenDesktop sites were spun, we had a way to prevent configuration drift. To the latest features for our customers, we bring also scheduled to update every 14 days to a newer version, and requires an easy way to push these updates. After all, we knew that we must have a robust disaster recovery solution.
Single Controller
to help all of these requirements, we establish the concept of a frame controller established. The core concept behind single-screen controller is that we will maintain a single base image for all our Delivery Controller. These base images are generated triggered as part of the product creation process of each code change. Once we near the beginning of our next release cycle, we have our QA team carry out a test, close to the next version. Once all tests are passed, we begin a gradual introduction of the new build for all customers.
We designed our operations so that each upgrade a new VM to rotate and connect to a database (which we regularly backup). Each machine failure or damage to the hard drive can be easily recovered by deploying a new Delivery Controller and it connects to the database - no different than any regular updating. This approach also prevents configuration drift (through reliance on durable VMs is avoided) and automatically wins the latest Windows security patches as part of regular upgrades.
baking frame controller and Deployment
As mentioned above, the goal is to have a completely generic base image. However, we have the ability to inject all tenant-specific data on the first boot time. While the base image jaws, we carry out the following tasks:
- Update latest Windows patches and dependencies
- Install to install the latest code XenDesktop product
- monitoring tools [configuring
- static VM
deploy to a controller from the base image, we have the following post-deployment operations to perform:
- customer configuration in injecting VM
- Connect to a dedicated database server
- create the VM XenDesktop website
, the goal is the amount of processing during the base image preparation as done to maximize this is a one time cost as with any tenant deployment / upgrade paid to a cost comparison. Moreover, by putting more in the base image, we get the test coverage and quality assurance. Our QA team can fully validate the base image before use in production. In this way, engineers can test to focus on important Runtime functionalities rather than on installing and static VM configuration.
Upgrade controller
We begin with a controller to version 1 with a matching XenDesktop database schema for version 1. This is a fully functional XenDesktop website is to be updated, ready.
We bring a new VM on a cluster, the publication on the next XenDesktop is code base. This cluster is based on the current frame image controller base.
It is important to note that the use of additional VMs simplifies the ability to take back a faulty upgrade. This approach is particularly profitable when. In a public cloud, where the cost of a few extra VMs less hosting is a lot for a short time compared to the purchase of additional hardware in a local data center
Before adding the new Controller cluster on the website, we take a backup of the database. This fuse provides us the opportunity to experience the initiate state back in case of problems during the upgrade.
We register the new controller cluster on the website. These controllers are expected to be backwards compatible with the previous database schema version.
Although the new controller is part of the site are, there is no DNS record to direct (CNAME) traffic on the new controller.
The database schema corresponds still the previous version. Prior to update the schema, we must quiescence to solve all controllers to ensure.
At this point, we have to update the database schema to the newer version.
[1945001[
After the database schema has been updated, we can restore the connection strings for all services on both the new and old controllers.
we also provide health checks on the site are available to ensure that the database upgrade was successful.
to show after a successful health check, we change our CNAME in DNS the new controller. This all VDAS will cause seamlessly to refer to the new controller, without a reconfiguration is required. We can then remove the previous controller from the website and delete the database backup.
At this point we have a successful upgrade performed, the user was seamless to the end.
treating disorders
in case of failure (at any point in the process described above), we can easily restore the upgrade rollback the database backup and remove the new controller.
If you are more interested in learning, there was a deep dive session of Citrix Synergy 2016 beginning of this year. For those interested in a study, please visit https://onboarding.cloud.com.
0 Komentar