Tag: cloud computing environments

  • 10 Actions IT Can Take to Curb Climate Impact

    10 Actions IT Can Take to Curb Climate Impact

    Digital technology is responsible for 3.7% of total CO emissions, according to estimates from The Shift Project. This could increase, as cryptocurrency mining is leaving a massive carbon footprint. With the world’s climate at stake, every percentage point matters.

    Thankfully, many tech companies have announced significant climate pledges to reduce emissions or offset their carbon footprint. For example, Microsoft states they will be carbon neutral by 2030. Amazon has committed to a deadline of 2040, and other companies such as Infosys have co-signed. Netflix’s climate commitment states they will reach net-zero greenhouse gas emissions by the end of 2022.

    Sustainability goals are vital long-term benchmarks, but on-the-ground technologists must make immediate optimizations to realize them. So, what exact steps can enterprise IT take to reduce their energy consumption and thereby reduce greenhouse gas emissions?

    I recently met with Anant Adya, Infosys, to explore the overall energy footprint of enterprise technology. Below, I’ll cover 10 ways IT can reduce its impact on the environment, including actions like auditing legacy infrastructure, optimizing cloud operations and moving toward renewable energy sources. Many of these energy optimizations bring the added benefits of cost savings and code consolidation.

    1. Reduce Resource-Heavy Legacy Infrastructure

    One potential cause of unnecessary energy consumption is legacy infrastructure. To reduce power consumption, Adya recommends going straight to a network compute storage setup. You may find energy consumption savings by porting resource-heavy legacy systems to the cloud. This will be especially effective if you’re running older end-of-life machines with less efficient processing.

    2. Reduce the Number of Clouds

    Next, look at the number of clouds your organization has adopted, Adya recommended. In recent years, companies have moved to multi-cloud setups for reasons like performance, cost or features. However, these movements could cause unnecessary duplication. For a large enterprise, refactoring your multi-cloud usage could mean reducing from five or six clouds to two or three, said Adya. With the right combination, you could very likely reach equivalent benefits with less reliance on multiple clouds.

    3. Reduce the Number of Core Data Centers

    Look at the number of sites in your core data centers and reduce the footprint, if possible. In his digital transformation work, Adya described working with a client to consolidate six large data centers into three essential locations — one for disaster recovery and two others to support American and European zones.

    4. Move Edge Locations to the Cloud

    Large organizations may be supporting a lot of edge infrastructure. Yet some locations, such as offices, don’t require that kind of infrastructure, Adya suggested. He recommended reducing local servers and embracing more cloud-based computing solutions, overall. He described one case in which an organization consolidated 600 edge sites to 350, thus reducing power consumption by 40%. Adya credited the cloud for delivering lower costs and greater scalability and resiliency.

    5. Embrace Cloud-Native Administrative Operations

    Cloud computing is arguably more energy efficient than traditional on-premises arrangements with its pay-per-use model and multi-tenant approach. To get the most out of the cloud, Adya recommended moving to the cloud, along with adopting cloud-native software. Most operational and administrative work can be processed in the cloud, he noted.

    6. Consolidate and Rationalize Resource Utilization

    Another way to stem energy consumption is to check for underused or forgotten assets. While the cloud can help optimize software, there is a good deal of unnecessary cloud spending on underused resources. In 2020, companies wasted $17.6 billion paying for idle resources and overprovisioned computing instances, reported ParkMyCloud. Adya thus recommended looking at the entire server and storage estate to calculate the server utilization and slimming down wherever possible.

    7. Increase Power Usage Effectiveness (PUE)

    Hardware is much more efficient these days — sometimes by as much as 10X more powerful with 10X less energy consumed. As storage technologies also are always improving, it could make sense to upgrade to new hardware to reduce overall power usage. It’s important to monitor power usage effectiveness (PUE) and check power ratings before and after migrations. Adapting your information life cycle management tiering practices for storage could also help optimize to reduce wasted power, Adya added.

    8. Meet Sustainability Goals on Campus

    Have you ever driven by an office park at night and seen thousands of empty rooms with all the lights on? This has got to change. Another way to reduce power is to limit a campus’s energy consumption, described Adya. Smart automation and remote monitoring could be used to manage power efficiency by turning off lights and heating and cooling to meeting rooms when empty, for example.

    9. Power the Business With Renewable Energy

    Run your campus with solar power. Adya described how Infosys is doing this by mounting 40MW solar panels at their corporate centers, in effect sourcing 43% of its energy from renewable sources. Another more passive way to invest in renewables is to choose a data center provider that commits, under contract, to using a high percentage of renewable energy.

    10. Purchase Carbon Offsets

    The previous goals are all active steps to reduce energy requirements. However, once an organization has attempted these optimizations, another option is carbon offsetting. This indirect method effectively cancels out a company’s emissions by funding green activities, such as tree planting.

    However, there is debate over the effectiveness of carbon offsetting, as trees take decades to become effective carbon consumers, and the practice doesn’t advocate for immediate change to harmful practices. “To be serious about tackling climate change, [organizations] need to stop carbon emissions from getting into the atmosphere in the first place,” wrote Alia al Ghussan for Greenpeace.

    IT’s Role In Curbing Climate Change

    Dire climate circumstances require quick action from all business units. In the process, organizations will likely consider many initiatives simultaneously to address their climate impact. However, as more companies become digital-first businesses, the onus falls on IT to support these changes.

    In addition to the above actions, there are undoubtedly other ways enterprise software architects can stimulate sustainability initiatives within their organization. This could include refactoring code into more lean iterations or eliminating unnecessary records storage. Tech leaders could also vocalize their support for global pledges, such as the Paris Climate Agreement, RE100 or Global Optimism.

    It would take a solar power plant covering 115,625 square miles, roughly the size of New Mexico, to power the world with solar energy! As you can see, it will take a lot of work to reverse our reliance on fossil fuels. But with scalable cloud computing and more intelligent infrastructure optimizations, IT can do its part to contribute to curbing climate change.

  • Accenture Cloud Simulation Platform Advances DevOps

    Accenture Cloud Simulation Platform Advances DevOps

    Accenture has launched a cloud simulation service through which DevOps teams can better determine which workloads will run best on different cloud computing environments.

    Kishore Durg, growth and strategy lead for cloud at Accenture Technology Services, said myNAV is a cloud service created by Accenture that leverages machine learning algorithms and other forms of artificial intelligence (AI) to leverage a knowledge base Accenture developed during its decade-plus working on cloud projects. DevOps teams can leverage that expertise without requiring Accenture personnel onsite to account for deployment variables such as technology infrastructure, types of application workloads, data, operational models and even business outcomes before committing to a specific cloud computing platform, he said.

    DevOps teams can also employ myNAV to engage in canary testing of application workloads, Durg added.

    DevOps teams employing myNAV are guided through a three-step process that starts with scanning and assessing the existing infrastructure, application and data landscape to recommend the most appropriate cloud architecture. An AI engine developed by Accenture then accesses a knowledge base curated by Accenture to come up with a recommended enterprise architecture. Finally, simulation and testing tools are applied to make sure the IT environment will scale as required.

    At the end of the process the DevOps team is presented with a bill of materials that resembles the bill of materials that might be presented when building a house, said Durg.

    Despite commodity pricing models, not all cloud computing platforms are created equal. Various attributes of application workloads lend themselves better to different attributes of cloud computing platforms. It’s not uncommon for IT organizations to find themselves rolling back applications from a cloud platform after encountering performance issues or discovering that cost of running a specific type of application on a cloud exceeded their available budget.

    Durg said 3-D visualization tools make it a lot easier to determine which cloud platform to employ and derive the maximum value from that investment by ensuring the cloud environment chosen is optimized for that application workload. Accenture research suggests that as many as two-thirds of organizations are not getting the maximum value from their cloud investments, he noted, and most of those issues stem from the cloud platform being poorly configured, added Durg.

    Whether by accident or deliberate, most IT organizations are employing multiple clouds to build and deploy applications, but their cloud platform selection process is often haphazard. Developers often default to the cloud platform that is easiest for them to access, which could create cloud platform lock-in. Cloud simulation platforms make it easier for the entire IT team to better appreciate the trade-offs in selecting one cloud platform over another.

    Of course, most IT teams would like to unify the management of multiple clouds under a single console through which they can more easily play one service provider off against another using a hybrid cloud computing model. Until that day arrives, however, most IT organizations will find it advantageous to know as much as possible about any cloud platform before making a decision they may come to regret.

    — Mike Vizard