Paige Vauter

One takeaway from DevOps Minneapolis Meetup: Charles Betz shared a principle which I didn't know about: treat platforms as products. For example, prioritizing what users actually need and bringing in people with a project management mindset or experience will significantly increase your ROI on building a new platform.

2

With #Couchbase, PG&E can connect its teams with this data, whether they’re offline or online, while improving service and lowering the cost of field visits. Cross datacenter replication (XDCR) adds resiliency, ensuring workers can depend on the application to be available when they arrive at the job site. Read their story

2

We’ve made warehouse provisioning and management in #Snowflake a lot simpler with Capital One Slingshot. Dynamic scheduling and pre-configured templates help you increase efficiency, streamline the provisioning process and reduce data spillage. See all the ways we’re doing #datamanagement right with Slingshot: https://bit.ly/41fUvn9

1

For anyone involved in large-scale data processing, particularly those considering Apache Flink or Google Cloud Dataflow. The article gives guideline on how to setup a fair benchmark consideration as well as potential optimization that can be done on the pipeline: Real-world Benchmark: It offers a practical, real-world benchmark study conducted by Yahoo, providing valuable insights into the cost and performance of these platforms under specific use cases. This can help organizations make informed decisions about their technology stack. Fair Comparison Methodology: The blog post outlines the methodology used to ensure a fair comparison between Flink and Dataflow, emphasizing the importance of controlling variables and establishing a consistent testing environment. This level of rigor adds credibility to the findings. Cost Optimization Focus: It focuses on cost-effectiveness, a critical factor for businesses. The post helps readers understand how to minimize costs on both platforms.

1

#AWS #reInvent2024 is here. Come join Tim Honychurch and I for an exciting session on new governance capabilities. We’ll dive into two impactful policies – Declarative policies, debuting at re:Invent, and Resource control policies. You will learn how to define and enforce configuration using Declarative policies and the significance of Resource control policies. 𝐄𝐯𝐞𝐧𝐭 𝐃𝐞𝐭𝐚𝐢𝐥𝐬:   💡 𝐓𝐨𝐩𝐢𝐜: COP378 New governance capabilities for multi-account environments  📅  𝐃𝐚𝐭𝐞 & 𝐓𝐢𝐦𝐞: Tuesday, December 3, 2024, from 5:30 to 6:30 PM   📍 𝐋𝐨𝐜𝐚𝐭𝐢𝐨𝐧: Caesars Forum | Level 1 | Summit 232 #reInvent #AWSOrganizations #EC2 #CloudGovernance #Security

36

1 Comment

🚀 What a week of announcements at AWS re:Invent 2024! Here are some launches I'm most excited about: 1. Amazon SageMaker Unified Studio: You asked and we delivered, an integrated experience for all our data and AI tools. Analytics and AI under one roof - this is going to streamline so many workflows. 2. Amazon Aurora DSQL & DynamoDB: Low-latency, multi-region consistency for both SQL and NoSQL. This is a game-changer for global applications. 3. Amazon Nova: The new suite of multi-modal AI models is impressive, the image and video generation capabilities are already adding value to my projects. Can't wait to explore more. 4. Amazon Q Developer: Automated code reviews, release management, documentation, and transformation capabilities. Q Developer is continuing to take the heavy lifting out of complex application migration and modernisation projects. 5. External Voice Transfer for Amazon Connect: Sometimes the oldies are the best. Voice transfer via SIP in Connect is going to make migrations so much smoother. 📅 Don't forget to watch out for the in re:Invent recap events in January. Hope to see you there. #AWSreInvent2024 #CloudComputing #AI #AmazonSagemaker #AmazonAurora #AmazonNova #AmazonQ #AmazonConnect #AWS #AWSNI

25

AWS announced the launching of RDS rightsize engine. Did anyone had a chance to review this? This is a long awaited feature that could bring a lot of value to customers in optimizing RDS costs, which tend to be significant for many organizations. Although not specifically mentioned in the article, I understand that recommendations will not be generated for instances covered with RI, which I find (very) odd IMHO. Did anyone had a chance to check this? #AWS great 1st step, looking forward to including Aurora in this engine, and more importantly provide a way to control RDS RI coverage, any way - convertible, SPs, marketplace. This would truly be a great holistic view to follow optimization best practice - rightsize than cover (then go do real FinOps). Great kudos to #FinOpsX for creating the stage for #FinOps announcements! https://lnkd.in/gNMSf-wk

14

2 Comments

I'm a little bummed. My team is headed to PGDay Chicago 2024, and I'm not going with them. The good news is Microsoft is a gold sponsor of PGDay Chicago which means there will be a booth, with socks, and stickers, and an activity book for Postgres too. Stop by, let the team know what you are doing with Postgres, and get some goodies.   And, don't miss the talks my team mates are giving: Arda Aytekin will share his "Learnings from Extension Development in Rust & PGRX" Claire Giordano will provide a "Map of Amazing Postgres Extensions You Might Not Know About" Krishnakumar R can be found "Probing PostgreSQL and Linux Kernel Interactions" And, Melih Mutlu will tell you about his "Takeaways from Hacking on Postgres as a Newcomer" #PgdayCHI2024 #PostgreSQL #Community #OpenSource #conference #Microsoft

24

4 Comments

Burnt ends and DevOps, it must be Kansas City! Last week I was honored to join DevOpsDays Kansas City for some great conversations. The open spaces are always my favorite part of andy DevOpsDays and there were great conversations on AI, technical documentation, and even salaries and compensation. I was thrilled to present two talks starting off with A Decade of DORA. The video of which is available now! Check out the talk and let me know how DORA has impacted your team over the past decade (give or take a five years). https://lnkd.in/eFR7nwKt #devopsdays #GBGB DevOps Research and Assessment (DORA)

60

5 Comments

🎉 @New Relic AI monitoring, the industry’s first APM for AI, is GA! New Relic AI monitoring empowers engineers with visibility and insights across the AI stack, making it easier to troubleshoot and optimize AI apps for performance and cost—all while ensuring data privacy and security. Get started:

17

1 Comment

🎉 @New Relic AI monitoring, the industry’s first APM for AI, is GA! New Relic AI monitoring empowers engineers with visibility and insights across the AI stack, making it easier to troubleshoot and optimize AI apps for performance and cost—all while ensuring data privacy and security. Get started:

35

1 Comment

So what's new in your #Copilot quest? A few recent highlights for me include joining an internal "Copilot Champs" Team, which has been fun, and just busy enough. If you are using Copilot at your place of work and are interested in helping to drive uptake, this kind of team can be a great resource. It's open to "civilians" like me, for whom Copilot is an interesting new tool to experiment with, but as you might imagine, at Microsoft there's a strong presence of experts who are part of the Copilot product team. Great info to be had. I also attended an optional training on using Microsoft Designer, which leverages DALL-E on the backend but offers some quick and easy ways to create formats like cool avatars, greeting cards, and coloring pages. It has a menu of styles that you can choose from to get closer to your target vision with fewer iterations. Although I will say that when I tried the avatar option, the result looked nothing at all like the photo I loaded. Similar to public use of DALL-E and chatGPT, you can use it for free for a limited number of prompts. Best #CopilotTip: DALL-E and most chatbots cannot understand or act on a negative command. So if you try to iterate on a result by saying "take the tree out of the background" it can't understand you or execute what you want. Try an alternative such as, "use a background with only grass" and see if that gets you closer to what you want. Given that I've been crowdsourcing a bit, for this post I thought I'd share a few public resources that I've found. (LinkedIn doesn't seem to do embedded links very cleanly, so find all links at the end of the post.) First, there's a great article from TJ Devine on getting more strategic with your use of Copilot. Second, variations of a fun prompt have been going my org around this week-- what kind of superhero are you at work? Check out the youtube link for the specific prompt. Lastly - connect to Designer and see how the specific tasks are a bit easier to use than going straight to DALL-E. https://lnkd.in/em_rB8iQ https://lnkd.in/eksVusSE https://lnkd.in/edxW9kQs

3

📢 💥 Attention AWS customers! If you are interested in #CloudGovernance, #CloudOperations, and #CloudSecurity topics and #AWSBestPractices, don't miss this #reInvent2024 session "COP378 New governance capabilities for multi-account environments" tomorrow from 5:30 to 6:30 PM at Caesars Forum | Level 1 | Summit 232. During this session, you will: 🔍 Deep dive into 2 new #AWSOrganizations policies 🏆 Learn best practices straight from the experts 🛡️ Enhance your #CloudGovernance & #CloudSecurity Elevate your cloud game - be there! 🚀

20

Attended the preconference session today at the annual Kansas City Developer's Conference. Enjoyed the session by Nate Schutta about "Cloud Native Architecture". Nate was very knowledgeable about the subject. He shared many good and bad examples of cloud architecture as well as many relatable personal stories. Here are a few takeaways which barely scratch the surface of all Nate's helpful presentation. ● automation, automation, automation is needed; do not rely on human memory. One example was given of one developer who missed upgrading one server which cost one company millions of dollars. ● speed matters; frequent deployments allows for making faster adjustments, fixing problems sooner. ●"we are all technology companies today" regardless of the industry in which we work. ● the right culture is needed which is "one of the most ignored aspects of software." ● Avoid temptation to follow "resume-driven design" where a tech is chosen because it would look good on a resume. ● CI/CD designed well allows for "injection points" such as the ability to add automated security scans and automated tests Several best practices were shared. Some points that resonated me included: "'● there should be one code base in version control, multiple deploys ● explicitly define your dependencies ● configuration must be separate from code ● dev (and testing) environments should match prod" Talked about differences between SLIs vs SLOs vs SLAs ● the difference between added "9s" in SLOs (such as 99.99% vs. 99.999%) availability is very costly, but paying extra for higher availability is worth it for certain critical workflows. Example: shopping cart availability is very critical and can result in immediate loss of customers if it's down. Microservices ● Are not the solution for all as they comes with an added cost. ● Can be good for code that has frequent changes. Doing a heat map is one technique that can help find the most frequently changing code. ● Should be independently deployable and scalable. ● Aim for high cohesion, low coupling. ● Observability is critical on such factors such as latency, traffic level, errors, etc.. ● Being agile and having short life cycles are important - the longer and more time is invested into developing a system, the less likely it is to be able to change course. ● Design them for failures even if you think no failures are possible, conduct resiliency testing and chaos engineering ● Automated deployments can make A/B testing, canary testing, feature toggles, and blue/green testing more feasible, resulting in earlier feedback, stronger disaster recovery, etc.. AI ● One hallucination in AI that is causing extra security vulnerabilities happens when it invents dependencies to libraries that don't exist ●AI-generated code can result in bad code that still needs to be reviewed such as to ensure it is meeting customer needs. #KCDC #KCDC2024

11

Having worn different hats throughout my career—as a developer, a program manager, and a product manager—across diverse organizations of varying sizes and flavors, I’ve rarely seen true product ownership that spans from ideation all the way to impact. This is what I call 360° Product Stewardship—the responsibility of managing the entire product lifecycle, from inception to post-launch success. Sure, organizational structures might force you to let someone else drive certain aspects, but the key is maintaining stewardship over the entire process. In doing so, I believe you’ll see far better results in the long run. #E2EProductManagement #360ProductStewardship

14

Explore the path to interoperability in this in-depth conversation with Michael Armbrust, the original creator of #DeltaLake, and Ryan Blue, an original creator of #ApacheIceberg. For many organizations, struggling with which table format to choose delays lakehouse adoption and ultimately harms their data + AI strategy. But there is an easier way. This insightful discussion covers the evolution of open table formats, how Databricks is solving interoperability, data governance and more.

1

Want to figure out the best way to leverage your data with LLMs? Check out this guide to get started!

8

It is so important, yet still underrated. Is the human resource (#HR) enablement and #EmployeeFirst approach fading in the face of the #AIFirst approach? However, in contrast, the #startup community continues to practice this approach effectively due to their small size and the expertise of their resources.

3

“True delivery speaks for itself—there’s no need for over-promising when real results are being achieved.” This idea resonates deeply with me because I’ve been on both sides of the table. One time, I was tasked with a UI rewrite and decided to use Ember.js—an adventurous choice at the time, as it was barely known in Melbourne. Version 1.0 had just been released, and I was the only developer on the team. For a month, I focused on laying the foundation, but with no visible output to demo, nerves started kicking in. People worried the project might not deliver. By the first scheduled demo, Product was preparing to push back. I calmly announced, "It's ready." They reviewed the work just an hour before the meeting. When the customer finally saw it, their reaction? “Wwwoooww!” 😄 This experience reminded me that meaningful delivery requires trust, focus, and sometimes a bit of risk-taking. It’s all about delivering value, not just promises. #ContinuousDelivery #SoftwareEngineering #Delivery

3