Insights about Cloud Testing Implementation
Insights about Cloud Testing Implementation
#1 - Clarity is everything
There are always many new software products that need to interact seamlessly with each other. It’s best to note down the answers to these questions right away:
- Which products are being used and what purpose do they serve for the organization?
- How do these products communicate with each other?
- Where can behavior be adjusted/modified?
- What dependencies does a product have?
There are numerous documentation options. We recommend a healthy mix of text and graphics, preferably in UML notations such as structural and sequence diagrams.
PRO TIP
If UML is not possible? Then it’s better to partially sketch it in PowerPoint than not write anything down at all.
#2 - IT expertise is more important than domain expertise when it comes to the implementation
When transferring test tools to the cloud, only minimal knowledge about the test specimen, i.e., the System under Test, is required. What’s much more crucial is to have a fundamental understanding of IT. If lacking, this should be arranged.
#3 – start small, improve fast
At first glance, starting Cloud Testing with cost-effective Spot Instances may seem tempting.
However, our recommendation is: Start small. Ideally with a unit. This way, you’ll learn how everything fits together and can expand with this knowledge. In the end, having detailed knowledge of the individual aspects speeds up the process. Unfortunately, there are no shortcuts when it comes to learning.
Learnings Cloud Kosten
The purely financial perspective on the question ‘Is Cloud worth it?’ is difficult to answer. However, we are definitely of the opinion: YES.
In particular, the switch to the Cloud is worthwhile if there is already consideration of acquiring a dedicated data center. The operating costs are at least comparable and in most cases even lower.
Moving to the Cloud significantly reduces work compared to having your own data center, and it can scale infinitely. Hardware failures are not a concern. We also assess the security of established Cloud providers as very high.
The main costs lie in usage. With a well-thought-out operational strategy, you can strike a good balance, and the costs can be estimated well over a manageable period of time.
Financial Benefits
- No initial investment costs
- No waiting time for the provision of a data center
- No need for 120 percent hardware capacity for compensating load peaks
- No hardware failures due to nearly inexhaustible scaling by the Cloud provider
- No subsequent investments for expansion and in case of repairs
- No personnel costs for hardware operation and maintenance
- No need to keep personnel on standby for emergency cases
- No electricity costs
- Arbitrary scaling through On-Demand possible
- High security standards (Security)
- Always up-to-date software in operation
- Always compliant-conformant license management (at least for the portions of the Cloud providers)
When is a cloud solution not worthwhile?
- If huge amounts of data are generated every day and need to be available in the cloud (hybrid model)
- If there is already a sufficiently large data center in place, including services
- If there is a lack of trust in cloud providers
- If there are no significant usage spikes in general
Conclusion
It's quick
Test results from days on-premises
in hours via the Cloud.
Cloud computing has the potential to significantly reduce the duration of long test runs. Our setup has demonstrated that the necessary tools are available and work out-of-the-box.
The cloud is cost-effective
Cheaper than On-Prem, scaling tests is almost a gift.
The additional costs for scaling tests can be adjusted through several levers.
In our view, tests in the cloud are significantly cheaper than in dedicated data centers.
The setup is easy
Few domain-specific knowledge required, but tool and IT expertise is essential.
The effort to set up a cloud testing infrastructure primarily requires tool and IT expertise.
We estimate the overall effort to set up testing in the cloud to be significantly lower than building dedicated data centers.
AWS Security Features
When implementing a CI solution in the cloud, data security is a crucial aspect. Thanks to the comprehensive security features provided by cloud providers, the cloud infrastructure can be protected. For example, AWS offers Identity and Access Management (IAM), Virtual Private Clouds (VPCs), and security groups for inbound and outbound traffic. IAM allows access control to cloud resources by assigning user roles with specific permissions. Security groups act as virtual firewalls that regulate inbound and outbound traffic of the cloud resources, restricting communication to authorized connections.
Cloud Security with AWS VPCs and Monitoring
AWS VPCs offer an isolated environment for the CI infrastructure, providing additional layers of security. Monitoring, logging, and analyzing cloud activities enable the detection of security-related events and timely responses.
In our case study, we examined the benefits of using the cloud for software testing in the automotive industry and described various use cases. Cloud technologies can shorten test execution times, accelerate feedback loops, and reduce costs for test infrastructure. The use of cloud-based solutions, especially in CI pipelines, provides development teams, IT, and management with efficient ways to optimize the testing processes, scale resources, and improve the time-to-market of software-based vehicle functions.