Top 6 Factors for Success in the Cloud
Migrating to the cloud isn’t as simple as some vendors propose. In fact, there are six factors that you need to take into consideration to ensure maximum efficiency and long-term success.
The first thing you need to do is estimate future budgets based on your workloads. To do so, look at your production use cases that show workloads at scale that represent real user queries, concurrency, user service level agreements, and query complexity.
Next, you should secure expert guidance to avoid common pitfalls. Migrating to the cloud is a complex process that can drive budget overruns. What worked on-premises may be very different in the cloud and migration is much more than just moving data. You also need to avoid unexpected costs while keeping your projects alive. Make sure you can scale without unexpected costs. As user demand changes, automatically scaling cluster size at the “power of 2” will deliver an unexpected bill that can kill your project—and your reputation.
Meet service level goals with software first. Cloud native data warehouses are designed for unlimited resources that automatically scale out using expensive compute resources to meet user demand. Mature software has free performance features in the cloud that can be optimized and tuned to meet service level goals without incurring the incremental cost of compute resources.
Demand flexible cloud pricing that offers both blended pricing (reserved) and consumption pricing (pay-as-you-go) that enables you to manage your costs and meet the needs of ever-changing user demand.
Finally, measure price-performance. Cost doesn’t equal price/performance. The metric you want to look for is the lowest cost per query. Vendors deliver multiple different metrics to make their solutions look the most attractive. Cost per query is accomplished by looking at an analysis of total system cost vs. number of queries executed daily.
The first thing you need to do is estimate future budgets based on your workloads. To do so, look at your production use cases that show workloads at scale that represent real user queries, concurrency, user service level agreements, and query complexity.
Next, you should secure expert guidance to avoid common pitfalls. Migrating to the cloud is a complex process that can drive budget overruns. What worked on-premises may be very different in the cloud and migration is much more than just moving data. You also need to avoid unexpected costs while keeping your projects alive. Make sure you can scale without unexpected costs. As user demand changes, automatically scaling cluster size at the “power of 2” will deliver an unexpected bill that can kill your project—and your reputation.
Meet service level goals with software first. Cloud native data warehouses are designed for unlimited resources that automatically scale out using expensive compute resources to meet user demand. Mature software has free performance features in the cloud that can be optimized and tuned to meet service level goals without incurring the incremental cost of compute resources.
Demand flexible cloud pricing that offers both blended pricing (reserved) and consumption pricing (pay-as-you-go) that enables you to manage your costs and meet the needs of ever-changing user demand.
Finally, measure price-performance. Cost doesn’t equal price/performance. The metric you want to look for is the lowest cost per query. Vendors deliver multiple different metrics to make their solutions look the most attractive. Cost per query is accomplished by looking at an analysis of total system cost vs. number of queries executed daily.