Smarajit Dhar

Ever wondered what happens when cloud systems fail? Dive into the extraordinary story of UniSuper. On a seemingly ordinary day, UniSuper’s Private Cloud subscription was accidentally deleted due to an “inadvertent misconfiguration” during the provisioning of UniSuper’s Private Cloud services. This wasn't just a simple case of data loss; it was a catastrophic event that led to the deletion of UniSuper’s account across multiple regions. Despite having data duplication in two different geographies as a protection against outages and loss, the deletion of UniSuper’s Private Cloud subscription caused deletion across both of these geographies. In the face of this disaster, UniSuper was able to restore services because they had backups in place with another cloud provider! This incident underscores the importance of cloud-to-cloud backup, where data in the cloud is backed up on another cloud. It offers an extra layer of protection and ease of use for businesses already using the cloud, serving the same purpose as off-site backup methods like tape or disk backup. UniSuper’s story serves as a stark reminder of the potential risks associated with cloud computing, even with precautions in place. It emphasizes the importance of having a robust cloud-to-cloud backup strategy to safeguard against such ‘one-of-a-kind’ occurrences. Because in the world of cloud computing, it seems, anything is possible! https://lnkd.in/gYhEx-qX --- Curious for more insights on cloud computing? Subscribe to my newsletter for regular updates and analysis. https://www.zilhaz.com #CloudComputing #GCP #CloudBackup #DataManagement #DisasterRecovery #BusinessContinuity

5

Introducing Arinco's AI Build Centre -- a brand new initiative aiming to accelerate AI adoption at scale for businesses across ANZ. Let me explain why Arinco's AI Build Centre is different. Combined power -- Arinco bring together the power of people, processes and technology to set the right AI foundations and ensure long-term successful adoption (no more POVs that go nowhere). Arinco's own AI application/assistant -- The AI team have been busy developing 'Coco' a multi-skilled AI assistant which is able to be rapidly customised to specific customer needs - aiming to leap customers from POV to implementation and beyond. Proven approach -- We can't forget about our existing AI Done Right™ approach which has already proven successful .. and won many awards might I add 😉. So to pull it together -- What are the components of Arinco's AI Build Centre that help to accelerate our customers from POV to implementation?   🤝 PEOPLE: Our dedicated crew of AI, Apps and Data experts, working hand-in-hand with customer teams. 👉 PROCESS: Our proven AI Done Right™ approach to AI development and adoption. 💡 TECHNOLOGY: Leveraging Arinco's Coco application, customers can rapidly prove value and start building custom AI skills in a matter of days instead of months. Learn more or get started at: https://lnkd.in/gK-nnKYS #ArincoAIBuildCentre #AIInnovation #MSPartner #AzureAI #GenAI

14

One of the biggest concerns I hear from clients is that they're worried about outages and disruptions when migrating to the cloud. We recently saw a big disruption in the services of one of Australia's largest superannuation funds, UniSuper. Unfortunately, it seems the cause was due to a misconfiguration in its Google Cloud services, which resulted in the deletion of UniSuper's Private Cloud subscription. Amazon Web Services has experienced similar outages in the past, sometimes across multiple regions. Sadly, outages and misconfigurations are commonplace in the IT world, which is reliant on complex coding and configuration. However, there is a reason for optimism when it comes to cloud management and migration for companies. Even this case, Google's quick response and partnership enabled them to recover all of the core systems and data. UniSuper also had backups with another service provider, a model that other companies can also consider following to increase resilience.📈 The greatest level of resilience doesn't always come with the most agility, but how many companies can afford to lose all their data because of an error, or a service outage? Some top recommendations I make around cloud resilience and disaster recovery (DR) include: 1. 𝐄𝐯𝐚𝐥𝐮𝐚𝐭𝐞 𝐲𝐨𝐮𝐫 𝐝𝐢𝐬𝐚𝐬𝐭𝐞𝐫 𝐫𝐞𝐜𝐨𝐯𝐞𝐫𝐲 𝐬𝐭𝐫𝐚𝐭𝐞𝐠𝐲 𝐧𝐨𝐭 𝐨𝐧𝐥𝐲 𝐢𝐧 𝐭𝐡𝐞 𝐜𝐨𝐧𝐭𝐞𝐱𝐭 𝐨𝐟 𝐝𝐚𝐭𝐚, 𝐛𝐮𝐭 𝐚𝐥𝐬𝐨 𝐭𝐡𝐞 𝐰𝐡𝐨𝐥𝐞 𝐨𝐟 𝐲𝐨𝐮𝐫 𝐜𝐨𝐝𝐞 𝐚𝐧𝐝 𝐜𝐨𝐧𝐟𝐢𝐠𝐮𝐫𝐚𝐭𝐢𝐨𝐧. Data backups are all very well, but if you don't have your code backed up, it will take a lot longer to rebuild your systems. 2. 𝐀𝐮𝐭𝐨𝐦𝐚𝐭𝐞 𝐰𝐡𝐚𝐭 𝐲𝐨𝐮 𝐜𝐚𝐧 𝐚𝐧𝐝 𝐭𝐞𝐬𝐭 𝐭𝐡𝐨𝐬𝐞 𝐚𝐮𝐭𝐨𝐦𝐚𝐭𝐢𝐨𝐧𝐬 𝐚𝐭 𝐫𝐞𝐠𝐮𝐥𝐚𝐫 𝐢𝐧𝐭𝐞𝐫𝐯𝐚𝐥𝐬. This helps prevent costly human errors, and testing automations regularly in a controlled environment will help avoid misconfiguration issues. 3. 𝐌𝐚𝐤𝐞 𝐚𝐫𝐜𝐡𝐢𝐭𝐞𝐜𝐭𝐮𝐫𝐚𝐥 𝐜𝐡𝐨𝐢𝐜𝐞𝐬 𝐭𝐡𝐚𝐭 𝐩𝐫𝐢𝐨𝐫𝐢𝐭𝐢𝐳𝐞 𝐚𝐧 𝐚𝐛𝐢𝐥𝐢𝐭𝐲 𝐭𝐨 𝐦𝐨𝐯𝐞 𝐟𝐫𝐨𝐦 𝐨𝐧𝐞 𝐜𝐥𝐨𝐮𝐝 𝐩𝐫𝐨𝐯𝐢𝐝𝐞𝐫 𝐭𝐨 𝐚𝐧𝐨𝐭𝐡𝐞𝐫 𝐩𝐫𝐨𝐯𝐢𝐝𝐞𝐫 𝐚𝐬 𝐫𝐞𝐪𝐮𝐢𝐫𝐞𝐝. The more cloud services you rely on from a single provider, the harder it may be to find another provider that can replicate all of those services and configurations. Some things can be kept on-premise and backed up so you can recreate yourself in another cloud as required, and you should also prioritize an Infrastructure as Code architecture that lets you install your entire infrastructure where you will if you need to move. 4. 𝐏𝐥𝐚𝐧 𝐚𝐬 𝐭𝐡𝐨𝐮𝐠𝐡 𝐝𝐢𝐬𝐚𝐬𝐭𝐞𝐫𝐬 𝐰𝐢𝐥𝐥 𝐡𝐚𝐩𝐩𝐞𝐧. The reality is that all cloud services rely on code, and code is complex. There are many different ways for errors and issues to creep in. More via The Guardian 👉🏼 https://lnkd.in/gU-zuCaN How have you prepared for disaster recovery in the past? Are there strategies I missed? Let's discuss.

47

14 Comments

Hey Melbourne! I'm excited to announce the launch of Observability Think Tank - a regular meetup designed create an inclusive, collaborative space where industry professionals, innovators, and enthusiasts can come together to explore the latest in observability. If we’ve met, you probably know that I learn new concepts best by immersing myself in community and conversations. I was looking for a regular, dedicated forum to explore common challenges and hot topics in observability but I couldn't find exactly what I was looking for. Over the past three months I've had 50 conversations with observability leaders and enthusiasts (that's a lot of ☕ coffee and 🍺) and one thing became clear - this is a sector that is RAPIDLY innovating. 🔔 To make this meetup group as engaging and valuable as possible, I’m seeking your suggestions: - Topics for Discussion: What areas of observability are you most interested in? - Potential Guest Speakers: Know someone who’d be a great speaker or willing to volunteer? - Other Advice/Areas of Interest: Any additional ideas or suggestions you think would benefit the group? Check out our Meetup page to get involved! Hope to see you on October 17th for our launch! H 🙂 https://lnkd.in/gSWCeCzC #Observability #TechCommunity #MelbourneTech #Networking #TechMeetup #Innovation #Technology #ObservabilityThinkTank

62

3 Comments