Brian Kimani

Demystifying Oracle's Dynamic Performance Views (V$ Views) As an Oracle DBA or developer, understanding the real-time heartbeat of your database instance is crucial for performance tuning, troubleshooting, and ensuring smooth operations. That's where Dynamic Performance Views, commonly known as V$ views, come into play! These incredibly powerful views are continuously updated throughout the life of your Oracle instance, providing a wealth of critical information. Why are they so important? Real-time Monitoring: Get live insights into database activity. Performance Tuning: Identify bottlenecks, active sessions, and resource usage. Troubleshooting: Quickly diagnose issues like locks, long-running transactions, or instance status. Some Essential V$ Views Every Oracle Pro Should Know: V$PARAMETER: All about your initialization parameters. V$DATABASE: Database status, ID, creation date, and more. V$INSTANCE: Instance name, server, and current status. V$SESSION: Details on all active user and background sessions. V$LOCK: Crucial for identifying and resolving database locks. V$TRANSACTION: Information on current active transactions. V$LOGFILE, V$LOG, V$ARCHIVED_LOG: Your go-to for redo and archive log details. Mastering these views is a game-changer for effective database management. Dive in, query them, and uncover the secrets of your Oracle environment! What are your go-to V$ views for daily tasks? Share in the comments! #Oracle #DBA #Database #PerformanceTuning #OracleDatabase #SQL #VViews #TechTips

18

Most ServiceNow devs find debugging ACLs hard. Here's how I've learned to do it after 3 years of being a Sr. SN Platform Engineer. 𝟭. 𝗦𝘁𝗮𝗿𝘁 𝘄𝗶𝘁𝗵 𝗔𝗰𝗰𝗲𝘀𝘀 𝗔𝗻𝗮𝗹𝘆𝘇𝗲𝗿 This new-ish tool answers the question: What does this user actually have access to — and why? It evaluates a user, role, or group against a specific resource and shows pass/deny/skip status for read, write, delete. You see which layer denied access — business rules, ACLs, or security filters — without blind guessing. This should be your first step. It often shows you the culprit faster than searching session logs. 𝟮. 𝗜𝗺𝗽𝗲𝗿𝘀𝗼𝗻𝗮𝘁𝗲 + 𝗗𝗲𝗯𝘂𝗴 𝗦𝗲𝗰𝘂𝗿𝗶𝘁𝘆 𝗥𝘂𝗹𝗲𝘀 Once Access Analyzer confirms it's an ACL issue, turn on Debug Security Rules (System Security > Debugging > Debug Security Rules). Reproduce the issue while impersonating the user. The session log will show exact ACL evaluation output — which ACLs were checked, which passed, which failed, and why. This is the definitive "ground truth" for ACL logic. 𝟯. 𝗦𝗲𝘀𝘀𝗶𝗼𝗻 𝗟𝗼𝗴𝘀 𝘁𝗼 𝗖𝗼𝗻𝗳𝗶𝗿𝗺 Filter your debug output by table and operation (e.g., incident.read). You'll see timestamps and results for each evaluated ACL. This confirms execution order and the exact rule that failed. It's important to search the logs after you have a theory to test. Enable ACL debug first so logs just confirm your thinking. 𝟰. 𝗖𝗼𝗺𝗽𝗮𝗿𝗲 + 𝗦𝗶𝗺𝘂𝗹𝗮𝘁𝗲 If you still don't understand why one user can access something and another can't, use Access Analyzer's compare and simulate features. This will tell you which role or group changes would fix access — faster than trial-and-error in a dev instance. This step stops people from "permissions guesswork" (turning ACLs on / off, adding / removing users from groups and testing). You get visibility into what would actually work! If you're leveling up your ServiceNow skills, I run a free community called the ServiceNow Cert Study Gym — co-working sessions, study plans, and accountability. Link in comments. What's your go-to ACL debugging method? #ServiceNow #ACL #Debugging

25

1 Comment

Reflecting on my recent discussion with a successful senior leader from a traditional software delivery background, where I was advocating for the need for evaluation engineers. When I defined what they needed to do, he mentioned, " ok, we needed testers." To simplify, I said yes, we need testers, and the subsequent reaction was, 'Why am I saying evaluation engineer?' I could see visible confusion, and I said I am happy calling them testers; however, I wasn't convinced, as my mind was racing about finding the right talent who can understand the need and ask. In the context of AI solutions, the primary difference is in scope and approach: a Quality Assurance (QA) Engineer, aka tester, focuses on the overall process and system integration to prevent defects, while an Evaluation Engineer concentrates specifically on the AI model's performance, behaviour, and output quality using specialised AI metrics. I agree that traditional quality engineers can do the job; however, there is a learning curve. Anyone willing to show a desire to learn, be curious, experiment, and apply the proper techniques will get the task done, regardless of their title. I am keen to see what others think. Please feel free to comment, sharing your thoughts and experiences. Mibin Boban CMgr MCMI appreciate your thoughts, given that the Testing Framework was shaped and adopted with the AI models and cognitive layer in mind.

19

10 Comments

a great conversation with Upnishad Deo today at OpsTree Solutions on the importance of frontend observability. great discussion around client side rendering vs server side rendering of the webpage content, it's impact on observability and performance, understanding web vitals and balancing the need of user friendliness vs SEO needs. while it gets difficult to sit for a boring session even for a few mins, this discussion went on for more than 2 hours without even a single distraction or boring moment. Sandeep Rawat #observability #webvitals #browser #js #monitoring #CSR #SSR #RUM

21

5 Comments

Proposed Research: DBA-Centric Database Reliability Engineering (DBRE) My research, titled "A Formal Operational Intelligence Framework for Resilient Database Reliability Engineering in Hybrid Enterprise Systems," addresses the critical reliability gap that persists despite advancements in database automation. The core of my proposal is to shift reliability engineering from a reactive, process-driven approach to a formal, predictive, and risk-aware algorithmic model. The work is grounded in empirical evidence, drawing upon a unique, anonymized dataset of over 40 recurring production incidents (including persistent RAC node evictions due to I/O stalls, GoldenGate lag during schema evolution, and storage reclamation paradoxes). Key Research Contributions: I propose to make three original contributions to knowledge: 1. A Novel DBRE Paradigm: Formalizing DBA operational intelligence into a structured, four-layer framework (Observability à Intelligence à Decision à Governance). 2. Empirical Failure Taxonomy: Developing a rigorous, enterprise-grade taxonomy of failure patterns specific to semi-distributed commercial database architectures. 3. Formal Risk Modeling: Translating DBA troubleshooting logic into a formal algorithm that calculates an Operational Risk Score R(O) for any maintenance action, ensuring controlled execution and preventing configuration drift.

25

3 Comments

"Your Local is not Global" Samora Akinbulumo, CPM shared this during a soft skills session he facilitated for the Developers Foundry program, a Tech4Dev initiative, and it stayed with me. While it resonates strongly with developers, the message is relevant to anyone building a career. Samora spoke about the importance of storytelling, why engineers should take control of their narratives, both in the workplace and online. It’s about more than just technical skill. To grow into roles like Engineering Manager, Technical Lead, or VP, you need to be intentional about how you communicate your work and progress. Share your wins, even the ones that seem small. Bring others along in your learning journey. Talent and output are important, but so is the ability to connect, influence, and be seen. He also emphasized that competence is key. Building a personal brand shouldn’t come at the expense of your core skills. If people start to sense a disconnect between what you present and what you can actually do, it becomes difficult to sustain trust or opportunities. So it’s really about balance, being great at what you do and being able to show it clearly and authentically. What’s that one step you’re planning to take or already taking in shaping your personal brand? I’d love to hear.

9

2 Comments

Custom Agents are here for Amazon Q Developer. For me, they change everything. Let me break this feature down for you. What are custom agents? Agents in Q Developer are not a completely new concept. You might have used the pre-built dev agent, test agent, or review agent. All of them are specialized, autonomous AI systems within Q Developer designed to handle a special software development task end-to-end. Each agent is purpose-built for a specific set of workflows and tasks, shipped with tools that allow it to operate independently. With custom agents in Q Developer CLI, AWS gives us the power to configure our specialized agents. That introduces a new paradigm for working with Q Developer CLI and hopefully also other modalities as soon as they support this feature as well. Instead of relying on a generic opinionated assistant, custom agents let you precisely define which tasks and workflows it is responsible for. It allows you to control tools, permissions, and contextual resources that are available to the agent during a CLI session. All you have to do: write a JSON configuration file. At the core, an agent configuration file can include the following sections: - name — The name of the agent (optional, derived from the filename if not specified). - description — A description of the agent. - prompt — High-level context for the agent working similar to a system prompt. - mcpServers — The MCP servers the agent has access to. - tools — The tools available to the agent. These can be built-in tools from Q Developer or tools provided by MCP servers. - resources — Resources available to the agent, like existing documentation - hooks — Commands run at specific trigger points, either when the agent is initialized or when the user submits a message I am deeply impressed by this feature, and it already has proven value in my workflows for several use-cases: - A backlog agent to maintain my backlog of feature requests and bugs for promptz.dev - A project-intelligence agent to maintain my living documentation as input for other agents. - An operations agent with access to AWS services running my accounts for root cause analysis and mitigation of issues of an incident we had at DFL Digital Sports GmbH with our Bundesliga App.. Custom agent configuration files are stored as JSON files in specific directories: Global custom agents reside in "~/.aws/amazonq/cli-agents/{agent-name}.json" and with that being made available across all projects and directories on your system. Project-level custom agents can be added straight into your repository ".amazonq/cli-agents/{agent-name}.json" versioned via git, and with that being made accessible easily for your team.

123

10 Comments