NetDBA4U

An AI Agentic Polyglot Database and Infrastructure Monitoring platform. Sixteen autonomous background services — monitors, agents, and daemons — work together to monitor, analyze, detect, respond, and report across every database and host in your environment. Integrated AI Workbench for instant expert-level diagnostics.

Why AI-Driven Database Management?

The NetDBA4U platform replaces reactive, human-paced operations with autonomous intelligence that never sleeps.

<5s
Incident Detection
vs. minutes for human operators
5
Database Platforms
Oracle • PostgreSQL • MongoDB • MySQL • SQL Server
16
Autonomous Services
Monitors, agents & daemons — 24 × 7 × 365
0
Alert Fatigue
Severity routing — only the right people notified

Autonomous Fleet

NetDBA4U runs 16 autonomous background services — monitors that watch, agents that act, and daemons that orchestrate. Every service is a real, production-hardened Node.js process started by the platform.

🔍

Sentinel AI

sentinel-daemon.js — Cross-domain telemetry extraction from all database platforms. Collects top SQL, wait events, memory, storage, blocking locks, and replication status. Uses local LLM (Ollama) for anomaly detection; escalates to cloud AI when needed. Handles offline-event triage with SSH forensics.

🛠

Triage Agent

triage-agent.js — Autonomous incident response. Scans open incidents every 5 minutes, gathers SSH/SQL forensics and Data Guard telemetry, uses LLM to propose remediation commands, and queues them for HITL approval via cryptographically signed email links.

📊

Monitor Fleet

Eleven specialized monitors, each targeting a specific failure domain with configurable thresholds and auto-resolution: Ping (5s), Host (3min), Health, Scheduler (5min), AWR (daily), ORDS (5min), CPU (5min), Disk (30min), Storage (30min), Host Log (6hr), and HADR (60s).

📨

Report Agent

report-agent.js — Scheduled report execution via CRON-driven schedules loaded from the database. Generates HTML Morning Briefs and Monthly Audit reports using Ollama local inference. Delivers reports via email and stores them in the portal.

🔐

Agent Engine

agent-engine.js — Core execution engine powering the HITL approval pipeline. Validates every AI-generated command against OWASP safeguards before execution. Handles MCP script dispatch, action queue deduplication, and HMAC-signed token verification for email approval links.

🤖

AI Health Guard

ai-health-monitor.js — Probes AI provider availability every 15 minutes. Automatically shuts down Sentinel AI and Triage Agent when providers are offline, and restarts them when recovered. Prevents error loops and wasted API calls.

🌐

Portal Interface

Centralized web dashboard for managing database targets and OS hosts, viewing incidents, analyzing performance data, deploying schema changes, managing PDB lifecycles, AI-assisted diagnostics, multi-channel notifications, and configuring monitoring thresholds and alert policies.

What NetDBA4U Can Do

Polyglot Database Support

Monitor and manage Oracle, MongoDB, PostgreSQL, MySQL, and SQL Server from a single unified portal. Each database platform gets native-level monitoring with platform-specific metrics and diagnostics.

Oracle 19c / 21c / 23ai MongoDB 6.x / 7.x PostgreSQL 15+ MySQL 8.x SQL Server 2019+
NetDBA4U — Instances
Monitored Instances
🟣 ORCL19C / ORIONLXONLINE
🟦 POSTGRES / PGHOST01ONLINE
🟩 MONGODB / MGHOST01ONLINE
🟨 MSSQL2022 / WINHOSTLATENCY
🟣 PDBTEST / CDB01ONLINE

Oracle Multitenant Management

Full CDB/PDB lifecycle management directly from the portal. Create pluggable databases, build refreshable clones across CDBs, open/close PDBs, convert refresh copies to standalone, and deploy monitoring schemas — all with audit trails.

PDB Create Refreshable Clones Cross-CDB Cloning PDB Open / Close Schema Deployment
Oracle PDB Manager
CDB01 — Pluggable Databases
CDB$ROOTMOUNTED
  ↳ PDBPRODREAD WRITE
  ↳ PDBTEST (Refreshable)REFRESH DUE
  ↳ PDBDEVREAD WRITE
Create PDB Clone Refresh Open/Close

Real-Time Performance Analytics

Live dashboards with CPU utilization, I/O throughput, wait event breakdowns, SGA/PGA memory analysis, top SQL by elapsed time, active session management, and host-level performance metrics via SSH. SQL workbench with execution plan analysis and AI integration.

AWR Analysis Top SQL Top Sessions SQL Workbench Host Performance Wait Events I/O Metrics Memory Stats
Performance Overview — ORCL19C
Real-Time Metrics
CPU Usage34%
SGA Memory71%
I/O Throughput52%
Active Sessions18 / 300
DB Time Trend (1h)

Automated Alerting & Incident Management

Severity-based incident creation with automatic resolution detection, escalation policies, and full audit trails. The ping monitor detects unreachable targets within seconds and auto-resolves when connectivity returns.

Severity Levels Auto-Resolve Audit Trail Ping Monitor Host Ping Monitor
Incident Manager
Open Incidents
Tablespace USERS near full (92%)CRITICAL
MSSQL latency spike detectedHIGH
Runaway session SID=142 killedRESOLVED
Stale stats: 14 tables flaggedMEDIUM
🔔 Email • SMS • Teams • Slack alerts fired

Multi-Channel Notifications

Unified Communication Agent with severity-based routing across multiple channels. CRITICAL alerts hit every channel simultaneously while LOW alerts stay silent. Each channel can be independently enabled, tested, and configured from the portal.

Email (SMTP) SMS (Twilio) Microsoft Teams Slack Custom Webhooks Severity Routing
Notification Settings
Channel Routing by Severity
🔴 CRITICALEmail + SMS + Teams + Slack
🟠 HIGHEmail + Teams
🟡 MEDIUMEmail only
🟢 LOWSilent / Log only
✅ SMTP ✅ Twilio SMS ✅ Teams ✅ Slack

Host Infrastructure Monitoring

Monitor the OS layer beneath your databases. SSH-based host discovery auto-populates OS version, kernel, CPU, and memory details. Periodic TCP pings detect host outages with automatic incident creation. Real-time performance dashboards show live CPU, memory, disk I/O, and network metrics with process-level drill-down.

SSH Auto-Discovery TCP Ping Monitor CPU / Memory / Disk I/O Top Processes SAR / vmstat / iostat Blackout Windows
Host Performance — ORIONLX
OEL 8.x — Live System Stats
CPU (8 cores)28%
Memory (64 GB)61%
Disk I/O18%
Top Processes
ora_lgwr_orcl4.2% CPU
ora_dbw0_orcl2.8% CPU

AI Workbench & Workbench Integration

Built-in AI assistant for database diagnostics and remediation. Self-hosted deployments can run local LLMs (Gemma 4, DeepSeek, Qwen via Ollama) on your own GPU hardware for complete air-gapped operation, or use cloud AI with your own API keys. SaaS deployments use managed cloud AI (Gemini, OpenAI, Kimi) with fast, GPU-backed responses — no hardware investment required.

Analyze AWR reports, SQL execution plans, and host performance snapshots through conversational AI. Send any performance workbench report or host snapshot directly to the AI Workbench for instant expert-level analysis.

Self-Hosted: Local AI or BYOK SaaS: Managed Cloud AI AWR Report Analysis SQL Plan Review Host Snapshot Analysis Context-Aware Prompts
AI Workbench — Gemini
Analyze this AWR report—what's causing the high DB time?
🤖 The primary driver is db file sequential read waits (68% of DB time). Top SQL 8gkf3m9a5bh20 accounts for 42% of I/O. Recommend: review index access path and consider partitioning on the ORDERS table.
What indexes should I add?
📊 Based on the execution plan, a composite index on (CUSTOMER_ID, ORDER_DATE) would reduce I/O by ~60%…

Top Sessions & Active Session Management

Live view of active database sessions with real-time I/O and CPU metrics. Sort by I/O or CPU, drill into SQL execution plans, and take action — kill or cancel sessions directly from the portal. Auto-refresh keeps you watching the pulse of your databases.

Active Sessions I/O Sort CPU Sort Kill / Cancel SQL Drill-Down Auto-Refresh
Top Sessions — ORCL19C
Active Sessions (sorted by I/O)
SELECT a.order_id, b.customer_name FROM orders a JOIN...
SID: 142 • Physical Reads: 48,291 • CPU: 12%
UPDATE inventory SET qty = qty - :1 WHERE product_id = :2
SID: 89 • Physical Reads: 12,044 • CPU: 4%
CALL dbms_stats.gather_schema_stats('HR')
SID: 201 • Physical Reads: 8,110 • CPU: 2%
⛔ Kill SID 142 📋 SQL Plan

Enterprise Knowledge Base & AI Trend Analysis

NetDBA4U acts as its own autonomous documentation engine. Resolving incidents or wrapping up AI diagnostics natively writes the resolution data straight into a centralized, searchable Knowledge Base repository perfectly correlated by our Generative Models.

Generative Diagnostics AI Trend Correlation Zero-Config Archiving Incident Resolution Trace Global Search
Knowledge Base Dashboard
✨ AI Trend Synthesizer (Last 20 Resolutions)
Based on recent archives, PGHOST01 is experiencing habitual blocking locks due to missing composite constraints on the inventory pipeline. I recommend adding composite constraints and implementing range partitioning.
Archived Matrix
Incident: Tablespace ExpansionINCIDENT
AI Session: Lock Latency TriageAI_CHAT
General: PDB Cloning PolicyDOCUMENT

Full Security & Audit Logging

NetDBA4U implements targeted zero-trust database-level access controls. Every action taken within the portal by users or autonomous agents is comprehensively tracked, integrating directly with Oracle Unified Auditing to surface database-level DDL changes and authentication failures.

Portal Audit Trail Oracle Unified Auditing Zero-Trust Target Access Action Tracking Granular RBAC
Audit Log
Portal Activity Trail
admin • KILL SESSION 142DBA ACTION
operator1 • VIEW PERFORMANCEREAD
admin • DEPLOY SCHEMA v1.0.3DDL
agent • RESIZE TABLESPACE USERSAUTO-REMED
🔒 AES-256-GCM • Zero-Trust Credentials • RBAC

AI Agentic Security & OWASP Pipelines

We restrict Generative AI autonomy behind military-grade OWASP Safeguards. Our Human-in-the-Loop authorization matrix traps destructive workflows (DDL, Drops, Grants, Container Switches) generated by the engine. Additionally, all outbound Cloud LLM requests pass through our mandatory Comms Sanitizer middleware, actively stripping topology details, masking passwords, and embedding firm 'DO NOT TRAIN' model restrictions to eliminate data leakage.

OWASP Defensive Agenting Comms Sanitizer Middleware Comms Sanitizer Human-in-the-Loop Queue Destructive DDL Traps Approval Routing
Approvals Queue
Pending Administrator Validations
ALTER SESSION SET CONTAINER=ORCLPDB1;
Context Shift • Blocked by Engine
DROP TABLE HR.EMPLOYEES CASCADE CONSTRAINTS;
Destructive DDL • Blocked by Engine
✅ Validate Execution ⛔ Reject Trapped Payload

Built for Enterprise DBAs

📦

Schema Deployment Engine

Version-controlled schema deployment with manifest-driven execution, idempotent scripts, CDB/PDB awareness, and rollback tracking. Propagate packages across all targets in one command.

🛡

Encrypted Credentials

Per-target password storage with AES-256-GCM encryption. Global defaults with per-instance overrides. HMAC-signed tokens for lightweight polling endpoints. Credentials never exposed in logs.

📊

AWR Workload Analysis

Automated daily AWR analysis across all targets. Detects adverse conditions (buffer cache ratio, I/O waits, parse ratios, DB time anomalies), creates incidents, and routes reports to AI for expert analysis.

👥

Role-Based Access Control

Multi-user portal with ADMIN, OPERATOR, and READONLY roles. Database-level access restrictions, session management, AI token protection, and full login/action audit logging.

📄

Stale Statistics Detection

Automatic detection of tables and indexes with stale or missing optimizer statistics. Heuristic-based alerts with one-click statistics gathering and CSV export.

🔧

Self-Healing Capabilities

Configurable auto-remediation policies. The platform can automatically respond to known conditions like tablespace near-full, long-running queries, or archive log buildup.

📉

Dashboard Analytics

Real-time donut charts for database target connectivity, host uptime, and incident status breakdowns. Collapsible topology tree grouped by platform, host, and service.

📤

Data Export & Reporting

Export performance reports, incident histories, and telemetry data. Save workbench reports as HTML, export stale statistics as CSV, and generate on-demand AWR analysis reports.

🔌

Oracle Scheduler Integration

Monitor and trigger Oracle scheduler jobs from the portal. Automated collection every 5 minutes, daily purge, AWR analysis at 8 AM, and on-demand job execution with failure detection.

💾

Automated Backup & Patch Management

Scheduled RMAN, PostgreSQL, MySQL, and MongoDB backups with configurable retention policies. Encrypted OTA patch delivery via .netpatch packages with RSA-4096 signed manifests and automatic rollback on failure.

🌐

Cloud Webhook Integration

Native webhook targets for OCI Functions, AWS Lambda, and Azure Functions. Trigger cloud-native automation directly from incident workflows with cryptographically signed payloads and configurable retry policies.

Enterprise Security Architecture

Every layer of NetDBA4U is hardened for enterprise deployment. From MFA to encryption to zero-trust access — your data and credentials are protected.

🔐

DUO MFA

Universal Prompt MFA via DUO Security for every user account. Enforced second-factor authentication with push, SMS, or hardware token support.

🔒

AES-256-GCM Encryption

All credentials and AI chat content encrypted at rest with AES-256-GCM. Transparent encryption/decryption on database read/write with automatic key rotation support.

🔏

Zero-Trust SSH

Key-based host authentication only. Dedicated netdba4u OS user for target access. No passwords stored or transmitted. Full SSH command audit trails.

Session Security

Comprehensive session security with hardened cookie policies, automatic inactivity termination, and protection against session hijacking and fixation attacks.

🔑

Password Policy

Comprehensive and robust password policy with complexity enforcement, automatic rotation, history validation, and secure hashing. Configurable to align with your organizational standards.

👥

RBAC & Audit

READONLY, OPERATOR, and ADMIN roles with database-level restrictions. Every portal action and agent decision logged. Oracle Unified Auditing integration for database-level DDL and auth events.

Edge Agent — Push-Based Monitoring

The NetDBA4U Edge Agent sits inside your database network, queries targets locally using OS authentication (Oracle wallet, / as sysdba), and pushes encrypted telemetry outbound over HTTPS. No inbound firewall rules required.

Per-target cryptographically secure API keys with TLS 1.3 transport. Acknowledged push with target-side sync confirmation. Ideal for SaaS deployments, air-gapped networks, and environments where inbound access is prohibited.

OS Authentication TLS 1.3 Per-Target API Keys No Inbound Rules SaaS Ready Air-Gapped
Edge Agent — ORCLPDB1
Push Telemetry
Last Push2s ago
Targets12
StatusCONNECTED
TransportTLS 1.3
🔒 AES-256-GCM • Ed25519 API Key • Acknowledged

Six Agents. One Dashboard. Total Control.

All sixteen autonomous services report their status in real time. Every incident, every action, every metric — visible at a glance.

Agent Status Dashboard
Autonomous Fleet Status
🔍 Sentinel AIRUNNING • 5m sweep
🛠 Triage AgentRUNNING • 5m cycle
📨 Report AgentRUNNING • scheduled
📊 Ping MonitorRUNNING • 5s cycle
📉 AWR MonitorWAITING • 08:00
🔐 Agent EngineSTANDBY • ready
Agent Fleet Status
Dashboard — Summary
Platform Health
5/5
Instances Online
2
Open Incidents
4/4
Hosts Up
16/16
Services Active
Recent Activity
Stale stats gathered on HR schemaAUTO
AWR analysis sent to AI • 08:00DONE
Portal Dashboard
Top SQL — ORCL19C
Top SQL by Elapsed Time
SELECT o.*, c.name FROM orders o JOIN customers c...
Elapsed: 48.2s • Executions: 1,204 • Reads: 92K
UPDATE acct SET balance=balance-:1 WHERE acct_id=:2
Elapsed: 21.7s • Executions: 28,441 • Reads: 14K
SELECT COUNT(*) FROM audit_log WHERE action_dt > :d
Elapsed: 14.1s • Executions: 302 • Reads: 8.2K
📋 Explain Plan 🤖 Ask AI
SQL Performance Workbench

See NetDBA4U in Action

Request a demo or more information about how NetDBA4U can transform your database operations.

Request a Demo