Using serverless and managed services is critical for achieving big wins with small teams on tight deadlines in the cloud.
Choosing a cloud platform and fully embracing its capabilities is key to success, even though leveraging multiple platforms for different strengths can also be beneficial.
Serverless computing allows teams to focus on business outcomes, accelerating product launches, reducing team sizes, and shifting the focus to more differentiated work.
AWS Glue is a managed service for building ETL jobs on AWS, eliminating the need to manage server infrastructure and making it easy to implement analytics pipelines.
Automating the deployment process of Glue jobs with a CI/CD pipeline, using tools like GitHub Actions, can streamline the workflow and ensure continuous deployment of ETL processes.
Using GitHub Actions, you can convert Jupyter notebooks to Python scripts, upload them to S3, update Glue jobs, and configure AWS CLI for deployment, making the process efficient and scalable.
Conducting high quality retrospectives following incidents is crucial for improving incident response practices by examining real-life incidents.
Facilitate retrospectives effectively by designating a facilitator, ensuring representation from key participants, and maintaining a blameless environment for open discussions.
Creating incident postmortem documents after retrospectives helps in documenting key information, discussing causes, impacts, resolutions, and lessons learned for continuous improvement.
Designate an incident commander who is calm under pressure and can manage communication effectively. They play a crucial role in incident management.
Choose appropriate coordination mediums like a war room or chat room combined with a conference bridge to effectively communicate and coordinate resources during incidents.
Focus on understanding the problem, identifying solutions, and implementing them quickly and safely. Reviews and retrospectives are vital parts of the incident management process.
ETL processes are vital for data analytics, involving extracting, transforming, and loading data for storage in a warehouse.
GCP offers options like Data Fusion and Cloud Dataprep for implementing ETL pipelines, catering to varying technical skill levels and preferences.
Alternative approaches on GCP for ETL include using services like Cloud Dataflow for more code-heavy processes or leveraging BigQuery for ELT if your team is SQL-focused.
Google-Managed Certificates in GKE are provisioned, renewed, and managed by Google, simplifying HTTPS setup for your domain.
Identity-Aware Proxy (IAP) in GKE provides zero-trust security, allowing secure access to applications without a VPN based on user identity and context.
Combining GCLB, GCP-managed certificates, and IAP offers a robust solution for serving and securing internal applications in the cloud.
Serverless architecture on GCP allows for quick application development with minimal operational overhead, setting Google Cloud apart from other providers.
Implementing a zero-trust security model on GCP, especially with context-aware access, enhances security for applications and services.
Transitioning from perimeter-based security to a zero-trust model with tools like IAP and IAM Conditions Framework provides a more flexible and secure approach, even beyond GCP.
During times of high demand, clear communication and expectation setting with users is crucial for maintaining trust and loyalty.
When facing unprecedented traffic spikes, focus on core functionalities and implement strategies like load shedding and graceful degradation to ensure application reliability.
Effective cloud infrastructure management, such as leveraging multiple availability zones, aggressive caching, and setting limits for resources, is essential in providing a good user experience and handling fluctuating demands.
Many established businesses are striving to transform into technology companies through digital transformation, but face challenges such as low throughput and quality issues.
The key to successful digital transformation lies in shifting organizational culture to become technology product companies where empowered product teams work towards a shared vision.
Focusing on vision, strategy, and empowering product teams is crucial for companies to effectively scale throughput and quality, rather than being caught up in tactical responses or inefficient processes.
Observability involves capturing various signals like logs, metrics, and traces to ask questions of systems without knowing those questions in advance.
Challenges in observability can include agent fatigue due to multiple operational tools requiring unique agents, capacity anxiety with elastic microservice architectures, and the need for foresight in collecting necessary data.
Implementing an observability pipeline can help in capturing wide events, consolidating data collection, decoupling sources and sinks, normalizing data schemas, and routing data to various tools for better observability in systems.
In microservice architectures, the conversation shifts from traditional monitoring to observability due to the complexity of multiple services interacting dynamically.
In static monolithic architectures, monitoring is more straightforward with a single runtime and centralized telemetry.
Observability offers deeper insights into system behavior by exploring new discoveries after the fact, providing more context and a higher level of granularity compared to traditional monitoring.
Avoid manually changing settings in databases or running CLI tools on production servers - it's risky and can lead to costly mistakes.
Invest in building admin portals for applications to access diagnostic tools and configuration settings - it pays off with increased safety and reduced time-to-resolution.
Admin portals allow for enforcing business rules, masking sensitive data, and empowering support staff to handle diagnostics and repairs - leading to more efficient resolution of issues.
Regular password changes may not necessarily improve security and can lead to weaker password choices.
Relying on username/password credentials for system access is a security risk, and using keys in their place is more secure.
Implementing tightly scoped service accounts, limited-privilege credentials, and continuous auditing can enhance security and mitigate risks in IT systems.
When structuring environments, consider the trade-offs between shared and team-specific environments based on costs, benefits, and complexities.
Different environment types (like playground, development, staging, and production) serve distinct purposes in ensuring developer efficiency, code validation, security, and reliability.
Minimize the number of environments to reduce costs, improve integration practices, and optimize developer efficiency, while balancing factors like data sensitivity and operational costs.
Cloud Identity-Aware Proxy (Cloud IAP) enables authentication and authorization for applications in Google Cloud Platform (GCP) by requiring users to login with their Google account and have appropriate access roles.
Configuring Identity-Aware Proxy involves associating it with an App Engine application or HTTPS Load Balancer and adding service accounts for programmatic authentication.
Authenticating API consumers with Cloud IAP involves generating a JWT signed with service account credentials, exchanging it for a Google-signed OIDC token, and making authenticated requests by setting the bearer token in the Authorization header.
NewOps is an evolution of DevOps, focusing on Operations through a product lens to shift Ops teams from masters of production to enablers of production.
Developer Enablement empowers dev teams to control their own destiny, extending responsibilities beyond building products to include testing, security, deployment, and operation of systems.
In the world of Developer Enablement in the cloud, Operations teams take on more advisory roles, providing domain expertise, guidance, and ensuring dev teams consider key operational aspects early in the development process.
Real Kinetic uses a declining-engagement model emphasizing mentorship to help clients for long-term success and quick value delivery.
Their assessment process involves on-site visits and delivering documents like Assessment Summary, Key Risks, Next Steps and Path to Production, and Best Practices and Recommendations.
Real Kinetic transitions from an active weekly/biweekly involvement to a monthly/quarterly retainer-based model, aiming to make clients self-sufficient while providing continuous support.
Multi-cloud can create unnecessary constraints and distractions, costing more than it's worth.
Disaster recovery, vendor lock-in, and pricing are main reasons why multi-cloud is considered, but they may not always justify the strategy.
For some large enterprises or specific use cases like leveraging the strengths of different clouds, multi-cloud may make sense, but it shouldn't be the primary focus for most companies entering the cloud space.
Systems are now more distributed and dynamic due to the rise of cloud and containers, requiring new tools and practices to support them
Observability in modern cloud-native environments involves gathering data for granular insights and empowered debugging through structured logging, metrics, traces, and events
Building an observability pipeline helps decouple data collection from ingestion into various systems and allows flexibility to add or replace tools without major disruptions
Google Cloud Platform (GCP) and Amazon Web Services (AWS) have different philosophies: AWS is favored by ops engineers for its low-level control, while GCP appeals more to software engineers with its managed services approach.
GCP gets criticized for poor historical support, less service offerings than AWS, and opaque roadmaps but is praised for integration of core services and high-quality releases.
Billing and support differ between GCP and AWS: GCP offers simpler pricing models and is enhancing support services, while AWS has complex billing and a more hands-on support structure.
Ops is evolving due to cloud, automation, and importance of DevOps. The focus is on automation augmenting us for better software.
Operations is shifting to NewOps, bridging cloud infrastructure and product development. Software is the future of Ops.
Specialization and scaling in Ops involve empowering development teams through tooling, automation, and a product mindset. Focus on enabling developers to self-service with guardrails in place.
Teams working towards a shared vision are more efficient and deliver more value. Clear vision empowers team members to make better decisions independently.
Having a shared vision prevents teams from iterating aimlessly. It provides direction and purpose to the team's efforts.
Strategic planning with high-level vision and concrete goals is essential for project success. Combining agile for day-to-day tasks with strategic planning ensures consistent delivery of value.
In the world of cloud computing, PaaS (Platform as a Service) has lost some of its appeal due to concerns like vendor lock-in and limitations.
The cloud landscape is complex and evolving rapidly, with tools like Kubernetes and serverless reshaping how applications are developed and managed.
Major cloud providers are moving towards unbundling and rebundling PaaS components to offer the benefits of accelerated development while retaining flexibility.
Companies go through different phases like seed, startup, growth, and maturity where culture and capabilities play vital roles.
Culture is the invisible force that guides your organization while capabilities stem from resources, processes, and values.
Success in organizations transitions from being resource-driven to process and value-driven over time, and adapting processes to new challenges is key for sustained growth.
Objectives and Key Results (OKRs) provide a framework for high level planning and resource alignment, helping to negotiate and communicate priorities within and between departments/teams.
OKRs are strategic tools, not task lists, and work well alongside tactical project management methods like Agile. They focus on outcomes and business needs, empowering teams to choose the best solutions.
Developing and applying OKRs is an iterative process involving negotiation, cross-team debates, clear objectives, and grading criteria to drive results and continuous improvement.