Migration to the cloud is a strategic decision. It has the potential to deliver improvements in service delivery, cost, agility and user experience. But, how should you assess the results of a cloud migration? What should you measure and how can you assess return on investment? .
New Ways of Measuring
Analytics that work well in your own internal data centers may be difficult to replicate in a cloud environment. For example, it is difficult to get full visibility into the whole cloud environment when applications may be located in multiple locations or moved from server to server in real time to optimize loading.
Those factors are outside your direct control and can have a significant impact on performance. Fluctuating network conditions can also influence results.
That makes an awareness and understanding of the variables and limitations important when you come to assess results. You should take these factors into account when setting your objectives and your key metrics.
Set a Baseline
Your assessment should identify the pre-migration performance baseline and the level of improvement you want to achieve. Migration should enable you to match or exceed previous performance.
Gartner has pointed out that migration objectives should take account of two types of infrastructure – mode 1 and mode 2. Mode 1 infrastructures reflect application requirements of reliability and efficiency. Migration objectives typically focus on cost reduction, faster provisioning times and reduction of the burden on IT staff.
In mode 2 infrastructures, the emphasis is on speed and agility. Migration objectives will aim to speed new service delivery to meet business objectives, rapidly scale IT capabilities, and take advantage of new technologies.
Audit Existing Performance
The first stage is to take an audit of resources before migration, measure response times and other metrics, and create a benchmark. Automated monitoring tools can save time and reduce the risk of error during this phase.
The audit should include measurement of the average waiting time for applications to load, response times to user requests and optimum performance levels, together with any metrics specific to your organization.
Records of uptime and the number and length of outages are important. You also need to measure traffic patterns since application-related traffic will move from internal networks to cloud connections.
You can then compare performance after migration to ensure that you are meeting your objectives and that cloud providers are delivering against service level agreements.
Test Potential Cloud Performance
To compare potential performance before a migration goes live, you can use monitoring tools to assess the performance of different providers and network services, as well as user experience in different location.
When migration is complete and the service is live, you need to monitor performance continuously. Although it is not possible to monitor cloud providers’ internal operations, you can use metrics that impact user and customer experience as a basis for comparison. By taking network metrics into account as well, you can identify the source of any performance issues.
Take a Rounded View
In addition to assessing operational issues and user experience, you should also review results against your business objectives. Did migration increase agility, improve service delivery or help accelerate digital transformation? This helps give a rounded view of the migration’s results and provides a basis for measuring ROIHow to Assess the Results of a Cloud Migration.