Stop the games, embrace high availability


Accessing computing resources must feel like playing Monopoly in some organizations. If you can out-game fellow analysts or have more-than-a-little luck, you get your report when you need it. If you can’t, you don’t pass Go and have to wait your next turn in the queue . . . whenever that is.

Here’s the deal. Your jobs, while important, are fighting for resources (CPU, disk and network), and the system has to make the right choices to meet everyone’s needs despite growing demand. The system has to manage mixed workloads with multiple analytical applications. These environments can, and usually do, have multiple users with different priorities, but there has to be balanced resource allocation.

You want – no, you need –  the flexibility to define workloads, schedule jobs and assign priority levels based on the user or group, the time of day, the type of application and more. Resource allocation shouldn’t be about who shouts the loudest or who has the most clout.

Reduce disruptions

Okay, no one likes downtime and we all want to avoid disruptions, but there is real operational danger to not having access to critical applications and data, because productivity and, ultimately, profitability are at stake.

This high-availability requirement means consistently keeping systems operational and accessible. When an unexpected event inevitably happens, the system can shift processing and automatically resume on servers with available capacity to minimize downtime in a planned and cost-effective way.

With grid computing, you can:

  • Effectively respond to system failure and planned downtime.
  • Schedule regular backups.
  • Monitor server processes and resource availability.

Jack up performance

And, with these ever-increasing demands, you need a way to decouple applications from the infrastructure in order to scale cost-effectively and grow without affecting activities and users.

Adding to these demands are user requirements for higher throughput to improve speed and meet service-level agreements. Having the ability to divide large analytics jobs into smaller tasks that can be run in parallel on smaller servers is critical.

To learn more about how grid computing can help manage computing tasks better and faster, read Building an Enterprise Analytics Platform: The Beauty of SAS Grid Computing.


About Author

Jeff Alford

Principal Editor

Jeff is a Principal Editor on the Thought Leadership, Editorial and Content team at SAS He's a former journalist with more than 30 years of experience writing on a variety of topics and industries for companies in the high-tech sector. He has a master's degree in technical and professional writing and loves helping others improve their writing chops.

Comments are closed.

Back to Top