In a survey we conducted last year, we discovered that many enterprises overestimate their ability to optimize their hybrid cloud infrastructures. More than three-quarters of respondents gave themselves high marks for a range of capabilities. Here’s the exact breakdown of the number who rated their abilities a 4 or 5 out of 5:
- Analyze their public cloud bills: 85%
- Dynamically adjust spending as needs change: 82%
- Predict long-term spending: 81%
- Forecast usage patterns: 80%
- Understand cost allocations at a granular level: 79%
- Get real-time cost alerting: 79%
- Identify unused/underutilized or abandoned resources: 77%
This sounds like great news, except there’s a problem. When we asked about how easy it is to get a global view of cloud costs, 42% said that it takes some effort. Furthermore, it takes a lot of time—46% said it takes hours and another 40% report that it takes days, weeks, or even longer. Given the time and effort these companies are putting into managing their cloud costs, those confidence numbers seem … overconfident.
So how can you know if you’ve got cloud optimization under control or if you’re fooling yourself? Here are 8 warning signs that you have a cloud optimization problem.
#1: Your cloud bills are unpredictable and your spend is exceeding budget.
Because of the pay-for-what-you-use nature of the cloud, there’s bound to be variability in your monthly usage and, therefore, your monthly cloud invoices. When higher cloud bills are the result of increased demand that drives revenue, that’s not a bad thing. If, however, you are seeing large swings that you can’t account for, and you’re busting your budget without seeing value for that excess spend, it’s a troubling sign that you lack control over your cloud infrastructure.
#2: You’re using native tools from each of your cloud providers and have to manually correlate it all outside of those tools.
Most cloud service providers (CSPs) offer native tools to help you analyze and optimize your cloud costs. Most enterprises, however, use more than one CSP. If you’re using the native tools, you’re either keeping your cloud cost data within individual CSP siloes, which means you’re not getting a global picture of your cloud utilization, performance and cost, or you’re having to spend a lot of time stitching together disparate data to get that comprehensive view. Either way, you’re shortchanging yourself.
#3: You can’t prove that you’re only paying for what you need.
There’s a difference between paying for what you use, and paying for what you need. If you’re not paying attention, or if you’re not making the optimal choices, the delta between the two could result in a lot of unnecessary spend. And the more budget you’re consuming in needless cloud usage, the less budget is available for other critical areas.
#4: You don’t have an easy way to get workload rightsizing recommendations.
The array of configuration options can number in the hundreds of thousands—just for one CSP. And providers are constantly introducing new types, sizes, and generations. The choice is overwhelming and you simply can’t find the optimal configuration needle in that haystack without a way to reduce the list to a manageable number of candidates. You could try to create your own shortlist, but how will you be sure that the best possible options are included? You’ll likely end up “goodsizing” or, if you’re lucky, “bettersizing” your cloud instance, but if you miss the truly “rightsizing” option, you’ll leave money on the table, so to speak.
#5: You don’t know which applications are overprovisioned.
The beauty of the cloud is that you don’t need to overprovision the way you do on premises. When scaling requires you to acquire, stand up, and configure computing resources, you build in some headroom so you’re not constantly doing that work. Additionally, that overprovisioning is a capital expense that you pay for once whether you use it now or later. None of that effort is needed in the cloud, however, so while you may want to build in a little buffer from a risk perspective, overprovisioning is an operational expense which means that you’re paying (and repaying) for extra resources unnecessarily. If you can’t see where you are overprovisioned in the cloud and by how much, you can’t rein in any unneeded capacity and spend.
#6: You don’t know if you’re paying for unused, idle resources.
It’s very easy to turn up new resources in the cloud, which is great for agility. But it’s just as easy to forget to later turn off or adjust those resources. Even if you remember to terminate an instance, you may forget to also terminate the storage, which has just become unattached. And that’s just one example. Cloud deployments are littered with unused compute instances, storage on stopped instances, unattached storage blocks, and other idle resources that continue to rack up charges. One company had 1,152 unused resources—which is by no means unusual—and when they were able to identify and address them, it created an instant $250K savings. When it comes to your cloud spend, what you don’t know can hurt you.
#7: You end up overspending on unused cloud capacity to guarantee performance and availability.
Performance matters. For many applications, slowdowns can have a big impact on the user experience that translates to lost revenue and customer churn. This requires you to make smart workload placement and sizing decisions. But if you can’t measure and evaluate with precision, you’re stuck with the brute-force and expensive approach of simply throwing money at capacity.
#8: You bought an expensive cloud optimization tool and you’re not sure it’s creating enough savings to justify the cost.
If your cloud optimization tool isn’t delivering significant, demonstrable savings, you’re not getting your money’s worth. Period. In contrast, Virtana Cloud Cost Management can reduce your cloud costs by 10% or more just over the course of the free trial, and the solution pays for itself, usually 5-10X over, in the first year.
Get cloud cost optimization under control with Virtana Cloud Cost Management
Virtana Cloud Cost Management radically simplifies management of your hybrid cloud IT infrastructure by optimizing cost, capacity, and performance in real time, on an ongoing basis. Virtana CCM delivers deep insight into your cloud spend across multiple cloud providers so you can:
- Understand cloud costs through bill analysis capabilities and cost vs. utilization reports
- Automatically optimize cloud instances with rightsizing recommendations
- Identify unused compute instances, storage on stopped instances, and unattached storage blocks
- Easily track the amortized value of reservation discounts at the instance level and conduct what-if analysis to evaluate potential savings
David McNerney
Product Manager, Virtana
