devopsy.kinsta.cloud

Category: Enterprise DevOps

  • Is it Time to Reconsider CI?

    Is it Time to Reconsider CI?

    DevOps is old, in technology terms, at this point. Many of you have never developed under a waterfall and have no idea what tools like Visual Test were, and that’s actually very cool. While knowing other processes and tools does help make informed decisions, being immersed in the tools that matter today is more important. We have entire DevOps stacks built up that allow for the development, building, testing and, increasingly, deployment of apps in an agile environment.

    As always happens, market leaders and those who compete with them, complement them or serve a subset of the market have grown up and become standards. Indeed, application release orchestration (ARO) was a phrase coined in the early days of DevOps that fell away completely; I heard it recently and thought, “Oh, are we bringing that name back?” Vendors have come and gone; some merged into other products, and some are just plain gone. Again, that is a good thing as the market, its needs and the customer base are all mature. While it is terribly painful to be the user of a product that just goes away one day, the prevalence of open source in the CI/CD landscape does buffer that pain a bit. And the market gets better as tools fall away and new ones take their place.

    But sometimes, the movement is slow, just simply due to inertia. Most of us – and by us, I mean IT shops, and by most I mean a lot – use Jenkins. The number seems to be on a steady downward trend, and the tool’s maintainers have definitely taken steps to improve the parts that make organizations consider other solutions, but the trend is there.

    And I, personally, think that is a good thing. Stagnation is what made waterfall such a weakness in enterprise IT. Everyone did it, so that makes it good. Same is true for the old phrases we’ve all heard at least one variant of – “No one ever got fired for buying IBM”, or “No one ever got fired for buying Oracle”, “…Cisco”, etc. The obvious answer to those statements then, as now, was, “Well, maybe they should have.”

    The same is true for Jenkins. It is a great system that can do everything. And that’s part of the problem. If you have small projects or are a team using it standalone with light requirements that all revolve around getting a quality product out the door quickly, Jenkins just might be too much for you. In the early days, Jenkins wasn’t, but the plug-in architecture and years of competition and changing market needs have made it huge. And while there is arguably better documentation than any other CI, and inarguably the largest user base to ask questions of, it’s still weighty to use in place of a SaaS that comes pre-configured and you can just hand it a repository with build files in it.

    There are a lot of enterprises that need what Jenkins has grown into–it did not expand in a vacuum. But there are many, particularly new, small, agile orgs or standalone teams that just don’t.

    There are a lot of options out there, and it is worth taking stock of what you are trying to do with your build/test process, and build a list of things required, then look seriously at what’s available. For some, lightweight is good.

    Of course, as much as one can be in our technological age, I am a minimalist. I like my D&D to be a rules-light variant, my dev tools to have enough functionality but not bury me in extraneous stuff, laugh at Excel for including abstract functions and then selling to consumers, etc.

    At my company that does development, we do indeed use Jenkins. But our projects range from simple knock-it-out one-offs to complex systems with multiple servers and clients, so we feel it is best for that range of variation. And like many others, the investment to move old projects would be questionable without a driving need. Not that I get to do much development these days myself, but I do stay in it just enough to be dangerous. But we have asked the question and determined the answer. For now.

    If you have issues with Jenkins, look around. There are options. They are not for everyone, but your org isn’t everyone. Find the best tool, not the most convenient, and keep kicking rear.

  • Applying AI/ML to Continuous Testing

    Applying AI/ML to Continuous Testing

    Artificial intelligence (AI) and machine learning (ML) can play a transformative role across the software development lifecycle, with a special focus on enhancing continuous testing (CT). CT is especially critical in the context of continuous integration/continuous deployment (CI/CD) pipelines, where the need for speed and efficiency must be balanced with the demands for quality and security. AI/ML contributes by automating complex tasks, predicting potential issues before they occur and providing actionable insights, thereby reducing manual efforts and enabling more strategic use of human resources.
    Moreover, the application of AI/ML extends beyond mere automation of tests. It encompasses the capability to learn from data, adapt to new information, and improve over time. This ability is invaluable for identifying patterns, anticipating vulnerabilities and optimizing testing strategies. In quality assurance, AI-driven tools can predict areas most likely to fail and tailor testing efforts accordingly. In security, ML algorithms can detect anomalies that signify potential threats, while in operations, AI can enhance feedback mechanisms, leading to more resilient and responsive systems.

    AI/ML for Continuous Testing

    Here’s how AI technologies can be applied to reduce bottlenecks associated with testing activities across various testing activities:

    1. Requirements Analysis
    • Explanation: Ensures test scenarios align with business requirements and user needs.
    • Bottleneck: Misinterpretation or incomplete analysis can lead to inadequate test coverage.
    • AI/ML Solution: NLP can automate the extraction and interpretation of requirements, ensuring comprehensive and accurate test coverage.

    2. Test Strategy
    • Explanation: Outlines the testing approach, objectives, and resources.
    • Bottleneck: An unclear strategy may lead to inefficient testing efforts and resource allocation.
    • AI/ML Solution: AI can analyze historical data to suggest the most effective test strategies and predict resource needs.

    3. Test Plans
    • Explanation: Detailed documents guiding the testing process, timelines, and responsibilities.
    • Bottleneck: Inflexible plans can struggle to adapt to project changes, causing delays.
    • AI/ML Solution: Machine learning algorithms can suggest adjustments to test plans based on ongoing project developments and past outcomes.

    4. Test Cases
    • Explanation: Specific conditions under which a test is executed.
    • Bottleneck: Time-consuming development and significant effort to maintain.
    • AI/ML Solution: AI can automate the generation of test cases from requirements documents, improving efficiency and coverage.

    5. Test Scripts
    • Explanation: Automated scripts that execute test cases.
    • Bottleneck: Script development and maintenance can be resource-intensive.
    • AI/ML Solution: AI can generate and update test scripts based on changes in the application or test cases, reducing maintenance effort.

    6. Test Data
    • Explanation: Data sets used during testing to simulate real-world scenarios.
    • Bottleneck: Creating, managing, and maintaining accurate test data is challenging.
    • AI/ML Solution: AI can automate the generation and management of test data, ensuring relevance and variety.

    7. Test Environment
    • Explanation: The setup where testing is conducted mirrors production environments as closely as possible.
    • Bottleneck: Configuration and maintenance of test environments are complex.
    • AI/ML Solution: AI can predict and configure optimal test environments based on test requirements, reducing setup time.

    8. Coordination with Dependent Systems
    • Explanation: Ensuring the system under test interacts correctly with databases and other applications.
    • Bottleneck: Dependency management can cause delays.
    • AI/ML Solution: AI can automate the detection and resolution of integration issues, enhancing coordination efficiency.

    9. Test Environment Setup
    • Explanation: Configuring infrastructure and tools necessary for testing.
    • Bottleneck: Setup complexity and resource contention lead to delays.
    • AI/ML Solution: AI algorithms can optimize environment setup, automatically adjusting resources as needed.

    10. Test Campaign Setup
    • Explanation: Organizing and scheduling a series of test executions.
    • Bottleneck: Requires careful planning and can be hindered by resource limitations.
    • AI/ML Solution: AI can assist in scheduling and prioritizing test campaigns based on risk and impact analysis.

    11. Test Execution
    • Explanation: The process of running test cases and scripts, both automated and manual.
    • Bottleneck: Time-consuming, particularly for manual tests.
    • AI/ML Solution: AI can prioritize test execution and identify flaky tests, streamlining the process.

    12. Test Verdict Reporting
    • Explanation: Determining and reporting the outcome of test executions.
    • Bottleneck: Manual verdict determination can be slow.
    • AI/ML Solution: AI can automatically interpret test outcomes, speeding up reporting.

    13. Logging of Data
    • Explanation: Recording data relevant to the test for further analysis.
    • Bottleneck: Extensive data collection can overwhelm resources.
    • AI/ML Solution: AI can intelligently filter and log pertinent data, reducing noise.

    14. Test Result Analysis
    • Explanation: Analyzing test outcomes to identify defects and issues.
    • Bottleneck: Requires significant time and expertise.
    • AI/ML Solution: ML algorithms can quickly identify patterns and anomalies in test results, highlighting potential issues.

    15. Test Result Reporting
    • Explanation: Communicating findings to stakeholders.
    • Bottleneck: Compiling reports is time-intensive.
    • AI/ML Solution: Automated reporting tools powered by AI can generate insightful and comprehensive reports quickly.

    16. Waiting for Resource to Fix Failed Tests
    • Explanation: Downtime while awaiting fixes for identified issues.
    • Bottleneck: Halts testing progress.
    • AI/ML Solution: AI can predict which areas might fail and propose potential fixes

    Challenges

    Applying AI/ML to software testing activities offers numerous advantages but also introduces several challenges. Addressing these problems requires a combination of technical solutions, process adjustments, and cultural changes.

    1. Lack of Intuitive Understanding of the Application Being Tested
    • Problem: AI/ML models may not fully grasp the application’s context or the nuances of its functionalities, leading to less effective test scenarios.
    • Solution: Enhance AI models with richer contextual data and incorporate feedback loops where testers can refine and adjust AI-generated test cases. Employing techniques like reinforcement learning can also help AI models better understand application contexts over time.

    2. Repeatability and Consistency Between Test Sessions
    • Problem: AI-driven tests may generate different outputs for the same input over different sessions, complicating test consistency and traceability.
    • Solution: Implement versioning for AI models and their training data, ensuring consistency across test sessions. Use deterministic approaches in conjunction with AI to maintain a core of stable, repeatable tests.

    3. Lack of Understanding of the Tests Generated
    • Problem: Testers may find it challenging to understand or trust the rationale behind AI-generated test cases, impacting their ability to evaluate test outcomes effectively.
    • Solution: Incorporate explanations into AI/ML models to provide insights into their decision-making processes. Foster a culture of trust and understanding through education and transparency regarding how AI models operate.

    4. Test Coverage
    • Problem: There’s a risk that AI/ML might not adequately cover all testing scenarios, potentially missing critical defects.
    • Solution: Combine AI/ML with traditional testing methods to ensure comprehensive coverage. Regularly review and adjust the criteria used by AI/ML models to generate test cases, ensuring they align with evolving application features and risks.

    5. Compatibility with Different Test Tools
    • Problem: AI/ML models might not seamlessly integrate with existing testing tools and frameworks, limiting their utility.
    • Solution: Develop or use AI/ML solutions with extensive API support and integration capabilities. Work with tool vendors or contribute to open source projects to enhance compatibility.

    6. Acceptance by Teams
    • Problem: Testers and developers may be skeptical or resistant to AI-driven testing due to concerns about job displacement or mistrust in AI’s effectiveness.
    • Solution: Educate and involve teams in the development and implementation of AI/ML testing strategies. Demonstrate the value of AI/ML in augmenting their roles rather than replacing them, focusing on AI as a tool to tackle mundane tasks and allowing them to focus on more complex and rewarding work.

    7. Data Quality and Availability
    • Problem: AI/ML models require large amounts of high-quality data for training. Inadequate or poor-quality data can lead to ineffective testing.
    • Solution: Invest in data curation and generation strategies such as synthetic data creation to ensure models are well-trained.

    8. Continuous Learning and Adaptation
    • Problem: AI/ML models may become outdated as applications evolve.
    • Solution: Establish continuous learning mechanisms where models are regularly updated with new data and feedback, ensuring they remain relevant and effective.

    9. Ethical and Bias Considerations
    • Problem: AI/ML testing models might inherit or amplify biases present in their training data, leading to unfair or discriminatory outcomes.
    • Solution: Implement ethical guidelines and bias detection methodologies for AI/ML model development and use. Regularly audit models for biases and correct them as needed.

    By addressing these challenges with thoughtful strategies, organizations can maximize the benefits of AI/ML in testing activities while mitigating potential drawbacks, leading to more efficient, effective and trustworthy processes.

    Summary

    This blog explained use cases of artificial intelligence (AI) and machine learning (ML) within the realm of continuous testing of the software development life cycle. By harnessing the power of AI/ML, we’ve seen how organizations can significantly enhance their development processes, making them more efficient, secure and responsive to user needs. The exploration of specific AI/ML applications across various activities has provided a blueprint for automating and optimizing tasks that traditionally required significant manual effort.

  • Driving the DevOps Evolution: ArgoCD, Tekton and Seamless Migrations

    Driving the DevOps Evolution: ArgoCD, Tekton and Seamless Migrations

    The landscape of CI/CD frameworks has evolved over the last couple of years. From the early days of Flux and Jenkins to the current focus on ArgoCD, this journey reflects the industry’s momentum toward efficient and reliable deployments.

    In recent times, we have seen many modern software delivery platforms adopt ArgoCD or Flux for automating deployments within their ecosystems. However, as multi-cloud deployments and scalability still remain a challenge, we are seeing the industry shift toward more powerful Kubernetes-native frameworks like Tekton.

    Evolution of CI/CD Frameworks

    Flux: Paving the Way for GitOps

    Flux was one of the pioneers in promoting the GitOps methodology. It allowed developers to manage infrastructure and deployments declaratively through version-controlled manifests.

    GitOps, as introduced by Flux, emphasized the use of Git repositories as the single source of truth for configuration, making deployments more reliable and traceable.

    Jenkins: The CI/CD Workhorse

    With its extensibility and vast plugin ecosystem, Jenkins quickly became the go-to CI/CD solution for many organizations. While Jenkins was powerful, its complexity and challenges in maintaining pipelines led to the exploration of alternative solutions.

    Spinnaker: Scaling Deployments

    Spinnaker emerged as a powerful tool for continuous delivery with a focus on scalability and multi-cloud deployments.

    Its support for canary releases, blue-green deployments, and pipeline orchestration attracted organizations aiming for more sophisticated deployment strategies.

    ArgoCD: The Rise of GitOps

    ArgoCD brought a renewed focus on GitOps, building upon the foundation laid by Flux. It provided a declarative way to manage Kubernetes applications, ensuring that the desired state in the Git repository is continuously in sync with the cluster state.

    The adoption of ArgoCD has been widespread due to its simplicity, scalability and ability to synchronize with multiple clusters.

    ‘Modern’ DevOps Platforms: Riding the ArgoCD Wave

    Many ‘modern’ software delivery platforms have either incorporated or integrated separately with ArgoCD to attempt a complete software delivery approach. This integration enables users to adopt GitOps principles seamlessly within their existing workflows.

    Here’s a glimpse into what it looks like to work with ArgoCD on these platforms:

    1. First off, you will have to orchestrate CI separately, which leads to context switching and a non-unified CI/CD workflow. You will need to leverage plugins on most of these platforms and hop back and forth between exploring steps and adding them to your pipeline for a few others.
    2. For CD, it is essential for you to connect your ArgoCD instance by keying in the server URL. This requires a working understanding of ArgoCD and also a working setup of the same before even signing up to these DevOps platforms.
    3. ArgoCD has to be defined as an artifact source and its specific steps must be added as deployment steps while configuring a pipeline. Other platforms require you to include steps or commands within the YAML files to interact with ArgoCD.

    While GitOps automation can be achieved by integrating ArgoCD with DevOps platforms, it has to be noted that these are not native workflows. There will be challenges that crop up, like:

    1. Improper synchronisation between repository and the deployment state on these platforms could result in unintended deployments or configuration drift.
    2. GitOps practices, when coupled with certain platforms, may have limitations in handling multi-cloud or hybrid cloud deployments.
    3. The complexity increases with the size of the infrastructure. The larger the number of clusters and environments the greater the number of pipelines and messier the orchestration.
    4. The fact that ArgoCD needs to be set up and running before even subscribing to DevOps platforms defeats the purpose of having low-code/no-code DevOps.

    The Rise of Tekton

    As the industry looks for solutions to address the challenges posed by existing CI/CD frameworks, Tekton has emerged as a promising alternative. Prominent tech enthusiasts in the DevOps space, like CloudBees co-founder Sacha Labourey, and market momentum make it clear that there’s a significant number of organizations working on integrating Tekton within their platforms, just like we saw with ArgoCD a couple of years back.

    What’s amazing is we already had platforms like Ozone, which started working on making Tekton enterprise-ready almost three years back and have further evolved over the years.

    Tekton: Container-Native CI/CD

    Tekton is a Kubernetes-native framework for building CI/CD systems. It provides custom resources to define CI/CD pipelines as code, allowing for greater flexibility and extensibility.

    This is done by introducing the concept of tasks that run within a pipeline. These tasks are reusable and pre-defined, all of which can be viewed in the Tekton hub.

    What ArgoCD achieves for GitOps is done with just one of the hundreds of task templates that Tekton comes with.

    Too good to be true? Well, every rainbow has its rain! Migrating from one framework to another is a huge task, and that is precisely why many enterprises shy away from migration.

    Can LLMs Help Automate the Migration of Existing Pipelines to Tekton?

    If you look at the diagram below, the answer is a resounding yes. The model is trained based on existing pipelines and what their corresponding Tekton pipelines look like.

    The overarching objective of this workflow is to facilitate a seamless conversion process using large language models (LLM) for generative AI.

    Central to this approach is using the chat context and the strategic application of roles for LLM models to establish context. Additionally, sample pipelines are leveraged to train and fine-tune the model to define how the corresponding YAML files of all these platforms look for any given pipeline.

    Here’s a more detailed read presented by Ozone as an industry-first last year at KubeCon North America.

    Conclusion

    The evolution of CI/CD frameworks reflects the DevOps industry’s dynamic nature and the demand for efficient software delivery pipelines. From Flux and Jenkins to ArgoCD and Tekton, each framework addresses deployment challenges.

    ArgoCD gained traction based on its GitOps principles and Kubernetes integration, yet faces synchronization, multi-cloud support and scalability hurdles. Tekton, a Kubernetes-native CI/CD framework, offers scalability, event-driven architecture and cloud-native practices to make up for these shortcomings.

    Whether embracing ArgoCD or exploring Tekton, the goal is resilient, scalable pipelines driving innovation and time-to-market acceleration.

    To hear more about cloud-native topics, join the Cloud Native Computing Foundation, Techstrong Group and the entire cloud-native community in Paris, France at KubeCon+CloudNativeCon EU 2024 – March 19-22, 2024.

  • The Great Debate: The Future of Humans and Software in the Age of AI

    The Great Debate: The Future of Humans and Software in the Age of AI

    Listening to the two sides of the debate regarding how much AI will take over software development, delivery and support tasks and its impact on jobs is both exciting for technology and scary for humans. The debate goes something like this:

    AI Advocate: Welcome to the future! With AI’s capabilities today, we’re looking at a world where software applications can practically develop, deliver and support themselves. The need for human involvement is diminishing rapidly.
    AI Antagonist: That’s an overly optimistic view. You’re underestimating the value of human creativity, empathy and ethical judgment. There are areas where AI simply cannot replace the human touch.
    AI Advocate: Consider design and specification. AI can generate prototypes and suggest improvements. It’s about efficiency and leveraging AI to cut down on the grunt work.
    AI Antagonist: But who sets the vision? AI lacks the understanding of user needs and market demands that only humans possess. The strategic decision-making and nuanced understanding of user psychology are irreplaceable.
    AI Advocate: True, but AI’s role in supervision and training cannot be overlooked. It can manage its operations, needing only minimal human oversight to update data and rules.
    AI Antagonist: And yet, who ensures these AI systems are aligned with ethical standards and current practices? Humans are crucial for overseeing these systems, ensuring they operate as intended and do not deviate into unethical territory.
    AI Advocate: I’ll give you that. However, in quality assurance and testing, AI significantly reduces the human workload by automating many processes.
    AI Antagonist: Automation has its place, but human insight is vital for identifying complex issues and understanding user experience in a way AI cannot. There’s a depth to exploratory and usability testing that AI can’t reach.
    AI Advocate: Ethical and legal oversight might be your strongest point. AI does struggle with nuanced decisions in privacy, security and bias. I concede that humans are necessary to navigate these complex fields.
    AI Antagonist: And don’t forget about customer support. AI can handle routine queries, but complex problems and sensitive issues require human intervention. The personal touch is something AI can’t replicate.
    AI Advocate: Innovation is a collaborative effort, I admit. AI assists, but human creativity drives continuous improvement and adapts to market trends.

    What Are the Implications if the AI Advocate is Correct?

    If the AI advocate’s vision comes to fruition, the implications for humanity could be profoundly transformative. First, the automation of routine and complex tasks alike would herald a new era of efficiency and productivity, potentially liberating humans from the drudgery of monotonous work. This shift could lead to an increased focus on creative, strategic and interpersonal tasks that leverage human strengths and foster innovation. Moreover, the widespread adoption of AI in software development could drastically reduce the time and resources needed to bring new technologies to market, accelerating the pace of technological advancement and potentially solving critical challenges in fields like healthcare, education and environmental protection. However, this future also raises concerns about job displacement and the need for societal adjustments to manage the transition, emphasizing the importance of re-skilling and education to prepare the workforce for a new economic landscape dominated by AI and automation.

    What Are the Implications if the AI Antagonist is Correct?

    On the other hand, if the AI antagonist’s perspective holds, the implications underscore the indispensable value of human involvement in technology. This scenario suggests that despite AI’s advancements, the nuanced understanding, ethical judgment and creative problem-solving capabilities of humans remain crucial. It highlights the potential risks of over-reliance on AI, including ethical dilemmas, privacy concerns and the perpetuation of biases if not carefully overseen by human intelligence. The emphasis on human skills and creativity could lead to a more balanced approach to technological development, where AI is used as a tool to augment human capabilities rather than replace them. This future would prioritize the development of technologies that are not only innovative but also ethical, user-centric and socially responsible, ensuring that technological progress benefits all of society.

    A Balanced Future

    Ultimately, the truth likely lies somewhere in between these two viewpoints, suggesting a future where AI and human capabilities are intertwined. The implications of such a balanced approach would be a collaborative ecosystem where AI enhances human efforts, leading to unprecedented levels of productivity and innovation while also creating new opportunities for human workers. This future emphasizes the importance of adaptability, lifelong learning and ethical considerations as central pillars of technological advancement. It also highlights the need for policies and educational systems that support the workforce in transitioning to roles that leverage human strengths in creativity, empathy and strategic thinking, ensuring that the benefits of AI are distributed equitably across society.

    Image source: https://vecteezy_futuristic-technology-concept-mixed-media-innovations-data_7019122_167-3.jpg

  • Survey: Significant Shift in IaC Tool Preferences

    Survey: Significant Shift in IaC Tool Preferences

    A survey of 350 IT professionals conducted by Firefly, a provider of a platform for managing cloud computing environments, suggests usage of infrastructure-as-code (IaC) tools is evolving rapidly in the cloud computing era.

    For example, more than 40% of the respondents reported that, to some degree, they are already using OpenTofu, an open source IaC tool that was launched as a fork of the Terraform IaC in the wake of changes to the software licensing terms it is made available under by HashiCorp. Well over half said they planned to use OpenTofu in the future. That compares to 60% using Terraform today, but only just over 20% said they planned to use Terraform in the future.

    The licensing changes HashiCorp implemented only affect third-party entities within a third-party platform. However, the survey indicates the number of end users affected by that change is more substantial than anticipated. A full 56% of respondents described the changes made to the Terraform license as disruptive.

    The survey also finds there is a lot more diversity in terms of the types of IaC tools being employed, with both the IaC tools from Pulumi and the open source Crossplane framework being advanced under the auspices of the Cloud Native Computing Foundation (CNCF) also gaining traction. The report finds more than 40% of respondents are also using Pulumi, with well more than half planning to use that IaC tool in the future.

    Crossplane, meanwhile, is now being used by 40% of respondents, with 60% planning to adopt it.

    Conversely, the report also notes the number of organizations planning to migrate away from CloudFormation from Amazon Web Services (AWS) is rising. While well over 60% use AWS CloudFormation today, only a third plan to use it in the future.

    Cindy Blake, vice president of marketing for Firefly, said as organizations employ additional cloud platforms, IT teams are clearly looking for tools that make it easier to centralize the management of multiple computing platforms. That trend will only accelerate as more DevOps teams adopt platform engineering as a methodology for managing the building and deployment of applications at scale, she added.

    In fact, noted Blake, it’s already fairly common for IT teams to use two or more frameworks. The report finds 57% of respondents work for organizations that are using two or more frameworks. Overall, the report finds that more than 64% of respondents have also codified more than half of their cloud assets.

    The number of organizations employing IaC tools to programmatically manage infrastructure is only going to increase, said Blake. The report, for example, found that three-quarters of respondents either already are or plan to use IaC tools to configure software-as-a-service (SaaS) application environments. In the longer term, advances in artificial intelligence (AI) will make it simpler for organizations to achieve that goal. Open source tools such as AiAC, for example, already make it possible to leverage large language models (LLMs) to generate code. Those types of tools will eventually eliminate the need to master specific programming languages to provision IT infrastructure as code, noted Blake.

    It’s not clear how IaC tools within the context of DevOps workflow will evolve, but the one thing that is certain is that the number of types of tools being adopted has never been more diverse.