Yogendra Rampuria

Adopting microservices isn’t always a smooth journey and in fact, it often fails without the right mindset and practices. In this tech talk, Vincent Tran and Steve Lam will unpack the common pitfalls, anti-patterns, and hard-earned lessons from real-world implementations. Learn how to navigate failure effectively, avoid costly mistakes, and turn missteps into a roadmap for resilient, scalable systems. Don’t miss out! Sign up now at https://ter.li/fgquc4

13

𝗜𝘁'𝘀 𝗵𝗮𝗿𝗱 𝘁𝗼 𝘀𝗲𝗲 𝗖𝗹𝗼𝘂𝗱𝘀 𝗶𝗻 𝘁𝗵𝗲 𝗗𝗮𝗿𝗸 - 𝗪𝗵𝘆 Redis 𝗠𝘂𝗹𝘁𝗶-𝗖𝗹𝗼𝘂𝗱 𝗠𝗮𝘁𝘁𝗲𝗿𝘀 𝗪𝗵𝗲𝗻 𝘁𝗵𝗲 𝗖𝗹𝗼𝘂𝗱 𝗚𝗼𝗲𝘀 𝗗𝗮𝗿𝗸 When a major cloud goes down, so do the apps, sessions & experiences that depend on it. Outages cost revenue, trust & time. Cloud-native services like ElastiCache or Memorystore work — until their cloud doesn’t. They’re locked to a single provider, with no easy failover across regions or clouds. One outage can take everything with it. That’s why multi-cloud resilience is no longer optional. With Redis’s Active-Active architecture, data stays synchronized across clouds and regions — automatically. - Zero downtime - Seamless failover - Consistent, low-latency performance everywhere When uptime is mission-critical, Redis Multi-Cloud gives you the control & confidence to stay online — no matter who’s offline. Because “always on” should mean always ready for the unexpected. #Redis #MultiCloud #HighAvailability #Resilience #CloudComputing #DevOps https://lnkd.in/emwFT4Xb

83

Just published a new article on a tricky Kafka Connect issue where connectors show RUNNING but silently stop streaming : https://lnkd.in/gitkj696 I break down how a rebalance + JDBC shutdown failure can stall Debezium tasks across MariaDB, PostgreSQL, SQL Server, Oracle, etc., and why deleting/reposting the connector fixes it. If you work with CDC pipelines, this one’s worth a read. #KafkaConnect #Debezium #DataEngineering #CDC

18

LiveRamp transformed its architecture for real-time data with SingleStore. The Chief Architect, Kannan D.R., discussed migrating to a scalable, cloud-native platform—enabling secure processing and cost savings. 🎥 Watch the recording on demand: https://ow.ly/yZni50XgXC2

95

1 Comment

Managing PostgreSQL extensions on Kubernetes has always been a balancing act between flexibility and immutability. But two upcoming features are about to change that: extension_control_path and ImageVolumes. Together, they make Postgres extensions immutable and flexible—solving challenges like bloated images, limited options, and operational complexity. With these enhancements, we’re unlocking new possibilities: 🔷 Dynamic extension installation without rebuilding images 🔷 Minimal core database images for leaner operations 🔷 Easier testing and upgrades 🔷 Stronger security with strict immutability This is the next evolution of Postgres on Kubernetes, paving the way for innovation across AI, GIS, time-series databases, and more. Get the full story in this blog by EDB's Gabriele Bartolini: https://bit.ly/4caCVEC #EDBPostgresAI #JustSolveItWithPostgres #Kubernetes #CloudNativePG #K8s #PostgreSQL #OpenSource #DevOps #CloudNative

24

In February, we announced that Tengyu Ma and the Voyage team would be joining MongoDB. Voyage is a pioneer in state-of-the-art embedding and reranking models, and their models are trusted by industry leaders like Harvey, Replit, and Snowflake. Fast forward a bit and our latest MongoDB Engineering Blog post is a deep dive by Frederic Roma into how Voyage AI's embedding and re-ranking models are transforming information retrieval within MongoDB, enhancing accuracy and performance for AI-powered applications. Now MongoDB is integrating Voyage’s best-in-class models directly into the database to simplify architecture, reduce latency, and to set the foundation for continuous improvements in accuracy, performance, and cost efficiency. For more, check out Fred’s full post! https://lnkd.in/ePrViKKX

89

1 Comment

Great news for Spanner and BigQuery users. Datastream now provides a seamless, serverless bridge from Spanner’s transactional power to the high-performance analytical capabilities of BigQuery and BigLake. Get the most scalable and available database that runs Google with BigQuery and BigLake’s serverless AI-Native Lakehouse!

514

3 Comments

EDB Day Session — Part 4: Aurora to PostgreSQL — Convenience Is a Strategy… Until Control Becomes the Requirement. In the same EDB Day session (Dec 8, 2025), the conversation moved to a topic that’s quietly becoming mainstream: Aurora. And the path out of it. Not because Aurora “doesn’t work.” It works—often brilliantly. But because many teams eventually hit a moment where the question changes from: “Is it managed?” to “Is it manageable… on our terms?” That’s the leadership inflection point. When you live inside an AWS-managed box, you inherit a lot of convenience. But you also inherit constraints—around upgrade timing, patching control, operational visibility, extension flexibility, and platform standardization across environments. And the hard truth is this: Modernization isn’t only about moving databases. It’s about moving decision rights back into your platform. The Aurora-to-PostgreSQL conversation in the room wasn’t ideological. It was practical: - How do we keep reliability high while we transition? - How do we reduce change risk during the move? - How do we standardize operations across cloud and non-cloud estates? - How do we keep upgrades and patching from becoming “special events” again? My belief after many of these rooms: The goal isn’t to replace AWS management. The goal is to replace uncertainty. The best migrations don’t start with tools. They start with an operating model: - workload-aware change windows - repeatable upgrade/patch playbooks - monitoring that reduces false alerts and suggests next actions - a path that avoids downtime drama and limits blast radius This is where the story of this series converges: Part 1: operational trust Part 2: reachable releases Part 3: always-on at scale Part 4: control without chaos Next: how teams make patching, monitoring, and optimization feel routine—using guided operations so upgrades become uneventful. #PostgreSQL #EDB #CloudMigration #Aurora #PlatformEngineering #SRE

28

Vibe coding is a big game-changer, enabling people to use natural language to instruct machines without needing to know how to code, said Cecily Ng, Vice President and General Manager for ASEAN and Greater China at Databricks, in a conversation with CNBC’s Jon Fortt at East Tech West 2025. Whether you're a student exploring AI or a developer at a major bank, Ng says access is widening. “The barrier to entry is lower, and now, success depends on giving the right instructions.”

67

3 Comments

It all starts with data, clean data and in real time. Sharing Confluent’s latest ebook on how real-time data is driving real AI results. It covers why legacy systems are holding AI back and how a data streaming platform can unlock speed, scale, and smarter decisions. Download it here: https://lnkd.in/gxGxu3Ff

23

When I was in Singapore working for Cloudera, Confluent was both a competitor and a partner. Apache Kafka was the heart of every modern data platform stack — open-source, powerful, and the backbone of real-time data pipelines across banks, telcos, airlines, and public agencies. Back then, we competed fiercely in some accounts and collaborated deeply in others. But the one thing everyone agreed on was this: If your data isn’t real-time, your insights aren’t real. Fast forward to today — IBM acquiring Confluent is a full-circle moment for me. It shows how central data in motion has become in the age of generative and agentic AI. From my seat now, leading the IBM–Google Cloud partnership, this move is bigger than one acquisition. It’s a signal of where the industry is heading: ➡️ AI needs trusted, governed, real-time data — not yesterday’s batch jobs. ➡️ Open-source roots (Kafka) are now powering enterprise AI at global scale. ➡️ Ecosystems — not individual vendors — will define the next decade. And this is exactly where our work with Google Cloud becomes exciting. Customers running Vertex AI and Gemini already rely on Kafka-based streaming. IBM brings enterprise governance, security, and hybrid cloud depth. Confluent adds a real-time nervous system for data. Together, this opens up a new frontier for clients in financial services, public sector, healthcare, telco, and regulated industries: 💡 Real-time data + Trusted governance + Advanced AI = Enterprise-grade AI that actually works. https://lnkd.in/eQxEk_M6

36

Tiger Data (creators of TimescaleDB) is now on Azure Marketplace. Handling heavy telemetry or IoT data? 225x faster queries than managed Postgres, 95% compression, sub-10ms response times. It's Postgres. Your team already knows SQL and Postgres. Now you get purpose-built time-series features: compression, data tiering, incremental materialized views. It’s worth checking out these benchmarks for yourself if time-series is part of your stack: https://lnkd.in/gvBj5sFc

107

3 Comments