Emerging Enterprise Architecture

Hybrid Cloud

Subscribe to Hybrid Cloud: eMailAlertsEmail Alerts newslettersWeekly Newsletters
Get Hybrid Cloud: homepageHomepage mobileMobile rssRSS facebookFacebook twitterTwitter linkedinLinkedIn


Public, Private and Hybris Authors: John Katrick, Elizabeth White, AppNeta Blog, Kong Yang, Nishanth Kadiyala

Related Topics: Cloud Computing, Hybrid Cloud

Blog Post

When Hybrid Cloud Becomes Hybrid Fog By @CWvanOrman | @CloudExpo #Cloud

Hybrid cloud offer tremendous opportunities for companies to cost-effectively achieve the performance-at-scale

Hybrid cloud is an appealing infrastructure model. With hybrid cloud, you can keep hosting applications and data on your own premises where necessary, while taking advantage of cloud economics and elasticity as appropriate.

Unfortunately, IT organizations are often unprepared to manage application infrastructure that spans the on-premise data center and multiple clouds. The result: they achieve vastly superior infrastructure economics - but still can't consistently ensure a great customer experience.

Forecast: Low Visibility
Companies typically start experimenting with cloud after running applications on-premise for years. That means their IT staff is used to having full visibility into all aspects of their application infrastructure: CPU utilization, I/O redlining, process hangs, systems logs, etc.

Most of that visibility disappears when you run applications in the cloud. Sure, cloud providers give you some metrics around compute, storage, network, and middleware. Their native tools, however, don't give you nearly as much infrastructure intelligence as you've become accustomed to with on-premise IT.

Some cloud advocates gloss over this loss of infrastructure visibility. They argue that the cloud has virtually unlimited capacity - so if you run into an I/O problem, you can always just spin up more VMs.

They also make light of the infrastructure visibility issue by pointing out that enterprise IT hasn't always done a great job of safeguarding application performance even when it has had rich infrastructure metrics.

There's some truth to these assertions. But inadequate visibility into cloud infrastructure is still a problem. Here's why:

  1. You don't want to throw your wallet at every problem. In the cloud, additional infrastructure immediately translates into additional opex. Cloud providers are thus happy if your knee-jerk response to performance issues is to simply add capacity.
  2. You want to be proactive, not reactive. Infrastructure metrics are essential for heading off many potential problems. In fact, those metrics are particularly useful for making sure you spin up additional cloud capacity before application performance hits a bottleneck, rather than after the bottleneck has already hurt you.
  3. Not every problem can be fixed with more capacity. Sometimes I/O redlines because a process has gone chatty, not because demand is actually spiking. If you can't quickly pinpoint and analyze those types of events, you can't deliver a consistently excellent user experience - no matter how much cloud resourcing you buy.

The benefits of hybrid cloud can't be bought at the price of lost visibility. To win in the application economy, you need adaptive infrastructure and full visibility into that infrastructure.

Unified Management for Hybrid Cloud
IT leaders who understand the importance of hybrid cloud management can take several steps to ensure their ability to effectively and efficiently optimize application performance. These include:

  • Determine exactly what infrastructure metrics your cloud provider(s) deliver - and which ones they don't. Incorporate these management value-adds in your criteria for choosing a provider(s).
  • Have a plan for filling the gap between what your providers deliver and what you need. Test this plan (including any technology) before committing to a cloud service.
  • Make sure your solution is applicable to multiple cloud providers. This will enable you to leverage your investments across multiple providers and mitigate the risk of provider lock-in.
  • Figure out how to aggregate metrics from your data center and your cloud providers. This aggregation is especially important if applications will traverse multiple domains.

Hybrid cloud offer tremendous opportunities for companies to cost-effectively achieve the performance-at-scale required for success in the application economy. If you can't keep a watchful eye on your cloud, though, you'll spend more and get less. That's why unified management of hybrid cloud is a must-have.

More Stories By Cameron Van Orman

Cameron Van Orman is vice president for solution and product marketing and enterprise management at CA Technologies. He leads a team of 20 people who together devise strategy for the business unit and develop all aspects of product marketing. Cameron and his team also lead the way for early product adoption in enterprise management.

Comments (0)

Share your thoughts on this story.

Add your comment
You must be signed in to add a comment. Sign-in | Register

In accordance with our Comment Policy, we encourage comments that are on topic, relevant and to-the-point. We will remove comments that include profanity, personal attacks, racial slurs, threats of violence, or other inappropriate material that violates our Terms and Conditions, and will block users who make repeated violations. We ask all readers to expect diversity of opinion and to treat one another with dignity and respect.