Virtualized Deployments and Why This flexibility is Key to Future Media Workflows – Greg Shay — Telos Alliance

Presenters : Greg Shay - Telos Alliance

Year : 2022

dlp_document_download : Virtualized deployments: Why this flexibility is key to future media workflowsGreg Shay, CTOThe Telos Alliance Te l o s A l l i a n c e : O u r V i s i o n●Why IT Technologies: What is driving this shift in technology?●Today’s IT Solution: Enabling new solutions in broadcast●Telos Alliance: Enabling today and tomorrow’s solutionsThe Broadcast Facility Implemented using Modern Server Software TechnologyInformation Technology is the foundation... 1.Hardware is unreliable.2.Human labor is expensive to react, reroute, reconfigure and repair.3.Many copies of identical devices are less expensive than many different unique devices. Broadcast has all these same needs…Recall a sharp focus:Why did the IT industry evolve to virtualized software: Server based virtualized software solves these problems with:●On-the-fly, automated replacement of failures from a pool of identical resources.●Operations automation, less labor. ●Shorten fault response time and improve preparedness with automated recovery for the primary function.●Orchestration of resource pool use, and deployment as needed ●Streamlined spares●Improved operational efficiency in operations, configuration, maintenance & monitoringThe more mission critical the function is, (i.e. the airchain), the more these methods are needed. This is more than the flexibility of "anywhere". Hypervisors(Virtual Machines)How are virtualized software products deployed?There are two main server platforms for deployment:ContainerizedSoftware What is Docker?Docker is a lighter weight method of OS-Level Virtualization, that allows software to be deployed portably, quickly, and reliably, under the control of deployment software.Software running software = automation = less labor How does it work?•Relies on containers, which package everything needed for software to run. •Containers eliminate the overhead associated with a full virtual machine. •Multiple containers can be run on a single server with a single OS. Why Docker? Designed for facility deployments•Linux server software deployed as a Dockercontainer.•Can be run on premises or in the cloud (AWS, Google, Azure,…)•As many instances as wanted can be run.•If you prefer VM’s, Docker hosting containers can be run virtualized in a VM.Note: Setting up these systems has its own complexity and special knowledge requirements: "DevOps" and "Infrastructure as Code". Redundancy means ‘keeping the product function running ’.The general strategy for redundancy using software in containers is: •If a server stops, start up the required containers on another server. •The very design of containers themselves is to make this as reliable, quick and easy as possible.•Kubernetesis an orchestration system often used.Redundancy with containerized products: Server uniformity and flexible network connectivity:•Saving labor: Orchestration software controlling softwareOne person can supervise the operation of thousands of machines.•The work to be done can be moved away from failed hardware, using 1 for N redundancy.•Can use geographic zones far enough apart to isolate from disaster outages and failures. (e.g. AWS Multi-AZ availability zones)Economics: Standardized virtualization containers allow the use of pools of servers, without rigid central resource planning. The same container can run on premises, or in the cloud, and can move back and forth controlled by automation on demand as needed.This gives the flexibility to use the server resource that fits the economic equation best:•Seasonal or on-demand consumption of 3rd party (cloud) resources.•Fixed contract for a professionally manage resources, either on premises (hybrid cloud) or remote. •Fixed 24/7 365 use constructed and managed in house.Additional benefits: Deployment LocationsBroadcast facilities require a range of deployments… Workflow, economy, and time-of-flight latency all influence where to deploy. The virtualized/containerized automated transitions take time measured in handfuls of seconds, they are not instantaneous.There is a fundamental economic question that cannot be avoided: Can the broadcast or media workflow be defined as successful with a rare multi-second outage?If so, can take full advantage of the economics of the best in class IT industry high availability by automation.Some reality biting: If a rare outage of ‘seconds’ cannot be acceptable: Operations are going to cost more and consume more resources for the deployment of simultaneous redundant end to end flows.However, the high availability automation (everything previously talked about) is still critically essential for keeping all redundant paths up.More reality… Thank You