Will Velida

Remember, in setups with multiple sites and a single logical layer, they shouldn't be relied on for disaster recovery. In some cloud setups, deleting a cloud account could wipe out all resources linked to that account. It's unclear if the cloud provider's safeguards were overridden by the delete action. It's really important to evaluate the risks when creating a solution. While many customers move to the cloud to avoid time-consuming design efforts, understanding your requirements and setting up an architecture on top of any cloud is absolutely necessary. Otherwise, issues such as higher costs, reduced availability, and recoverability could impact your business. Whether it's a Private, Public, or Hybrid cloud setup, some folks think stretched clusters can be a good disaster recovery solution because they offer better availability during site subsystem outages, workload mobility, and other advantages. However, stretched clusters aren't actually suitable for disaster recovery. Put simply, stretched clusters involve mirroring data over two nodes in different locations, creating a logical layer from which the data is served. This logical layer becomes a single point of failure (SPOF), no matter how many nodes are serving behind it in a distributed way. Just think about corrupted mirrored data spanning two sites and backed by redundant storage nodes. Can stretched clusters really help with recovery? Make sure to stay on top of logical errors, understand the cloud risks and constraints, and architect your solutions effectively. https://lnkd.in/eMmq9wnx #vmware #google #vcdx #cloud #cloudarchitecture #cloudofferings #publiccloud #hybridcloud #privatecloud #socialmedia

11

1 Comment

Most new Azure regions are now being deployed as 3+0 regions, meaning the primary region has 3 AZs and no satellite region. This marks a shift from the 3+1 (Primary and Secondary) architecture used for Azure regions built between 2019 and 2022. It is crucial to understand that 3 AZ doesn't mean 3 datacenters, because an AZ has one or more, and sometimes over 15+ in one AZ cluster. Historically, Azure regions were paired to ensure data consistency and availability, especially for services like Azure Storage, Azure SQL Database, and Azure Cosmos DB. This practice dates back to before 2017, when Azure Availability Zones were introduced. However, the need for regional pairs has decreased over time, and many new regions are now being deployed without a regional pair. This change allows for more flexibility in region selection and helps address capacity issues in certain areas. In the future, you might start seeing the new Azure extended zones meeting those satellite requirements. (https://lnkd.in/d7YHpvzD) When it comes to high availability (HA) and disaster recovery (DR), customers should begin by identifying the failure modes they are concerned about. Geo-replication is not very helpful when a customer accidentally deletes data in a storage account, as the deletion will be geo-replicated within a few minutes. Instead, Object Replication or AzCopy Sync can be scheduled with multiple filters to address this issue. I have been advocating for this "any-to-any" for multiple Azure Products for a long time, and it is exciting to see this becoming a reality for selected Azure products. The community needs to keep pushing this, because feedback is super important! #AzureReliability #RemovingTechnicalDebt #BuildingforTomorrow

18

2 Comments

Tune into our next ep of New Breakpoint for step-by-step demos that show how AKS can help you build resilient, scalable cloud-native applications. Register now: https://lnkd.in/gJQpdAAX #MicrosoftAustralia #MicrosoftNZ #MSDeveloperANZ #MSAzureANZ

7

Did you know that the effective use of data and machine learning (ML) could help win gold medals? Explore how Swimming Australia—the governing body for swimming in Australia and an AWS customer—stood up a data lake to provide its coaches with a detailed analysis of each swimmer’s performance. Ultimately, this is a story about using data and ML to gain a meaningful advantage in one of the most competitive endeavors on the planet. If that’s what AWS did for this team, imagine what it can do for yours. Watch the original AWS video. https://lnkd.in/dFc7aitw

17

As promised, #OpenAI and #AIServices examples I talked about in this week's Brisbane Azure User Group (run by Dan and Damien) are ready to use: https://lnkd.in/gPkAXSA9 It helps you for a potential extension to dive deeper into real enterprise customer scenarios such as Azure OpenAI Using PTUs/TPMs With API Management: https://lnkd.in/gJzTmBkg Deploying Azure OpenAI and AI Services using Azure Verified Modules (#AVM) adheres to Microsoft's Secure Future Initiative: https://lnkd.in/gW4QTQqK #SFI #SecureByDefault #Azure #AI #AVM #IaC #Bicep #Terraform

70

2 Comments

Swimming Australia uses AWS Panorama for real-time computer vision at the edge. The solution integrates an app with Panorama, SQS, Kinesis Video Stream, and App Runner. #aws #awscloud #cloud #amazonkinesis #amazonsagemaker #amazonsimplequeueservicesqs #analytics #artificialintelligence #awsiotcore #awspanorama #broadcast #customersolutions #datascienceanalyticsformedia #directtoconsumerstreaming #internetofthings #kinesisvideostreams #mediaentertainment #mediasupplychainarchive #messaging #topposts

I just noticed the exciting news from our Instaclustr by NetApp Apache Kafka team (posted today, July 4th in Australia!) about our new Kafka tiered storage feature (public preview) here https://lnkd.in/g79MK_PJ This basically answers the age-old question of "Is Kafka a streaming platform or a database?" with "both" (at least for time series data) and will open up potentially a whole bunch of different use cases for streaming historic data (e.g. ML training and replaying data into new sink systems via Kafka connect) but I'm sure there are more! I was interested in the potential for improvements in Kafka client and cluster performance due to the optimisation of real-time and historic workloads and improvement to cluster operations (e.g. broker recovery operations)! Great work team! I'm sure there will be plenty of opportunities to try this out and do some performance/scalability experiments and even performance modelling/predictions in the future.

10

Learning the Cloud is becoming a must-have skill for everyone in Data! In under 90 minutes you'll learn about Azure Synapse Analytics, SQL Databases, ADF and more! https://lnkd.in/eG4SAze7

272

11 Comments

🌟 What an inspiring day at KubeDay Melbourne 2024! Yesterday was packed with insights, innovation, and great conversations. Here are my three key takeaways from the event: 1️⃣ Enterprise-Scale Kubernetes: It was fascinating to learn how large enterprises are driving innovation and scaling efficiently with Kubernetes, enabling teams to have more autonomy while ensuring governance. 2️⃣ Kubernetes Gateway API: It was interesting to learn how it is evolving to overcome the challenges associated with Kubernetes Ingress, providing better traffic routing and control at scale while facilitating the separation of duties. 3️⃣ Kubernetes as a Platform Framework: The next evolution in platform engineering involves using the Kubernetes API as the interface that allows teams to streamline operations and enable faster development. 💡 Overall, it was an excellent opportunity to view the world of Cloud-Native from different perspectives, stay updated on the latest Kubernetes trends, and meet like-minded professionals. #Kubernetes #KubeDay #CloudNative #DevOps #GitOps #Innovation

93

5 Comments

New IRAP assessment out including new services. Many Australian customers are looking to experiment with how generative AI applications can help them better serve the Australian public. Customers can use two of the newly assessed services—Amazon Bedrock and Amazon DataZone—to help align with their governance, sovereignty, and security requirements up to the PROTECTED level.

14

Azure Region Provider Comparison With the release of the New Zealand North Azure Region on the horizon, I wanted a way to measure what capabilities can be deployed in the new region, particularly in comparison to Australia East. To do so, I wrote a script that uses the Azure CLI to query the provider namespaces of each region, compare them side by side, and output the results into an HTML table. #azure #mvpbuzz #nznorth #provider #powershell https://lnkd.in/gn_dJmr4

48

2 Comments

I put together a summary of my key takeaways of the Sydney AWS Summit experience on the Cevo Australia blog. #aws #summit #sydney https://lnkd.in/gadNfqmP

34

1 Comment

🛡️ When the Going Gets Tough, the Tough Build k8sPulse! I recently found myself needing an emergency Kubernetes monitoring tool—one that could run with just kubectl access, especially in those moments when everything else seems to fail. This need gave me a crazy idea: why not create my own emergency Kubernetes dashboard? What started as a simple Bash script grew into something much more! 🤖 I transformed my emergency dashboard idea into a fully functional, easy-to-use CLI called k8sPulse. It’s built in Python, leveraging the native Kubernetes client and enhanced with Pandas and Matplotlib for beautiful insights. I even challenged myself with design — despite not having any background in UX/UI or color theory, I decided to design the entire product, including its modern homepage: https://lnkd.in/gaap6b2e. 🛠 Want to try it out? If you have your kubeconfig ready, just install it with pipx, then, in any directory: k8spulse This will generate a full report in a docs folder. And if you have an OpenAI API key, you can enable intelligent recommendations! 💡 Check it out, and I’d love to hear your thoughts! #k8sPulse #OpenSource #Kubernetes #Python #CLI #AI #EmergencyMonitoring #SelfTaughtDesign

12

How can your cloud work smarter, not harder? Are you grappling with the high costs and complexities of manual scaling for your Azure SQL Database? I've got the perfect solution for you! 🚀 In my latest article I've detailed a step-by-step tutorial on how to automate database scaling based on CPU usage. This will ensure your database always operates efficiently and cost-effectively.  🔗 Read the full article here: https://lnkd.in/dK5Z2wmg  #Azure #SQLDatabase  #Automation #Cloud 

9

🤯Naming is hard! Ever feel like your cloud resources are playing hide and seek because of wonky naming? I just dropped another blog post on wrangling those unruly Azure resource names using Azure Policy and some clever #Bicep tricks. It's like giving your cloud a mini-makeover - less chaos, more consistency. Curious? Give it a read and let me know if it helps tame your naming zoo! Link, like always, in the first comment! #Azure #CloudComputing #IaC #DevOps #AzurePolicy

23

22 Comments

Want to visualize your MongoDB data effortlessly? 📊 With Natural Language Mode in Atlas Charts, you can create sophisticated data visualizations by simply asking questions in plain English. Learn more on this latest enhancement announced at #MongoDBlocal Sydney. 👇 https://lnkd.in/gefjGvK5

12

In today’s fast-paced tech landscape, managing infrastructure as code (IaC) effectively is key to ensuring consistency and efficiency. There are a lot of IaC languages out there, but if you are purely working with Azure, then let me introduce you to Bicep. A powerful tool for Azure infrastructure provisioning that can be a game-changer for your organisation if you are looking to scale while maintaining high standards. One of the biggest advantages of using Bicep modules is the ability to centralise and standardise configurations across multiple projects. By storing these reusable modules in a container registry (such as Azure Container Registry), teams can easily access and deploy consistent infrastructure resources, ensuring best practices are followed without reinventing the wheel. Here’s why standardising Bicep modules in a central repository can be a game-changer for organisations: 🔹 Consistency Across Projects – Centralised Bicep modules mean every project can inherit the same infrastructure standards, helping teams avoid configuration drift and ensuring security and compliance requirements are met. 🔹 Ease of Scaling – With standardised modules in place, scaling becomes as simple as referencing a shared module, enabling teams to deploy complex infrastructures across multiple environments and projects with minimal effort. 🔹 Faster Time to Deployment – Teams can focus more on solution-building and less on creating repetitive infrastructure code. Pre-approved modules reduce the time to get environments up and running, significantly speeding up project timelines. 🔹 Enhanced Collaboration and Reduced Technical Debt – Centralised modules serve as a single source of truth, fostering collaboration among DevOps teams, and minimising technical debt by reducing duplicate or inconsistent infrastructure scripts. 🔹 Flexibility and Control – Containerised Bicep modules allow for versioning, so updates to modules can be rolled out in a controlled manner across projects, ensuring smooth transitions and testing before deploying changes at scale. By adopting standardised, centralised Bicep modules, organisations can not only scale more effectively but also empower their teams to deploy infrastructure that is consistent, secure, and aligned with best practices. At Arinco, we specialise in setting up these solutions to help businesses streamline their infrastructure and achieve reliable, scalable results. Feel free to reach out if you have any questions or want to know more about how we can help with you cloud journey! #Bicep #IaC #Azure #DevOps #Infrastructure #InfrastructureAsCode #CloudComputing #scalability #Arinco

21

Not used #AzureAIStudio before?? Take 10 mins to watch this video then go get access - you’ll have your own copilot built before you know it! #copilot #genai #azure #azureai

17

This is a massive milestone for GitLab; A massive thank you to all the Federal Government partnership. GitLab is in-process for FedRAMP. So proud of this moment, so proud of GitLab for achieving this. Introducing GitLab for Public Sector. https://lnkd.in/efzpBa4w A single platform to manage your complete Dev-Sec-Ops and more. Why would one want to manage more than a handful of tools to begin with? Now you don’t need to even manage one tool! Come see the future. Efficient, modern, complete, and now Secure!

97

1 Comment