Complete AI Training
Sign inGet my AI kit

Your job's AI kit

Get your AI kit

Tell us who you are and what you do. We show you your kit right away and email you the link: skills, prompts, AI agents, MCP servers and courses for your job.

500+ jobs ready, and we make a kit for any other job. No payment needed to look.

Share

AI agent for database administrators

Connection Pool Saturation Agent

Connection use under the limit during peak load, with pool settings backed by test results

AnalysisIT and developmentOperations
See how it worksGet access
Connection Pool Saturation Agent: what goes in, what the agent does and what you get

What it does

At peak hour users see random timeouts, and the real cause is that the database ran out of connections. This agent reads connection counts, wait events and each application's pool settings, and identifies which services hold the most connections, which hold them idle and which seem to leak. It builds a revised pool configuration and tests it under a load test in a non-production setup. After each test it re-measures wait times, errors and connection use, and tunes again if the results are not yet within target. When the numbers hold, it prepares the change for production with the evidence. A person approves the production setting. Edge case: a service holding connections open during a slow third-party call is flagged as a design issue, not just a pool size.

How it works

Follow the arrows from top to bottom. The orange dashed arrow is the loop: when a check fails, the agent goes back and tries again.

Start and resultWhat it doesA check on its own workWaits for your OKGoes back and retries
Yes, continueYes, continueApprovedNoNo 1 STARTS WHEN Saturation alert or planned peak 2 USES A TOOL Read connection counts, wait events and poolsettings 3 DOES Find which services hold, leak or wait forconnections 4 DOES Draft a revised pool size and timeout for eachservice 5 USES A TOOL Run a load test with the new settings in staging 6 CHECKS THE RESULT Does peak use stay under 80 percent with notimeouts? If not: adjust sizes based on which service waited andretest. Back to step 3. 7 USES A TOOL Re-measure after the test and compare with thebaseline 8 CHECKS THE RESULT Is any service leaking connections after the test? If not: flag the code path and keep the cap while theowner fixes it. Back to step 3. 9 YOU APPROVE DBA approves the production setting 10 RESULT Tuning report and settings change
Read the steps as a list
  1. Saturation alert or planned peak
  2. Read connection counts, wait events and pool settings
  3. Find which services hold, leak or wait for connections
  4. Draft a revised pool size and timeout for each service
  5. Run a load test with the new settings in staging
  6. Does peak use stay under 80 percent with no timeouts?If not: adjust sizes based on which service waited and retest. Back to step 3.
  7. Re-measure after the test and compare with the baseline
  8. Is any service leaking connections after the test?If not: flag the code path and keep the cap while the owner fixes it. Back to step 3.
  9. DBA approves the production settingThe agent waits here for your OK.
  10. Tuning report and settings change

How it decides

Pool sizes are lowered for idle holders and raised only for services that wait. A setting passes when peak use stays below 80 percent of the limit with no timeouts.

  • Target peak use below 80 percent of the connection limit
  • Lower pools for services with over 60 percent idle connections
  • Flag connections held longer than 30 seconds
  • Stop after 4 failed tuning rounds and escalate

Make it yours

Every agent is a starting point. You choose these settings for your own situation.

  • Peak use target (default 80 percent)
  • Idle share that triggers a cut (default 60 percent)
  • Services in scope
  • Load test profile

What keeps you in control

It always asks you first

  • Changing pool settings in production

Hard limits

  • Never changes production settings without approval
  • Tests only in staging

It stops when

  • Done: tests pass and the change is approved
  • Stop: staging cannot reproduce the peak

Set it up

We guide you through the set-up, step by step

Members get the full set-up guide for this agent. No technical skills needed: you copy, paste and upload.

10 minto set it up in your AI
5 AIsChatGPT, Claude, Copilot, Gemini, Grok
  • One set of instructions to paste into your AI, with the clicks for ChatGPT, Claude, Microsoft 365 Copilot, Gemini and Grok
  • The agent then walks you through connecting your own data, one source at a time
  • A downloadable copy with the flow chart, the rules and the full guide
Get access to this agent

An example run

What happensDuring the Friday peak, 480 of 500 connections were used. The agent found the checkout service held 220, mostly idle, and the search service waited on 30 timeouts. It cut checkout from 220 to 120 and raised search from 40 to 70. The first test showed 85 percent use, so it adjusted to 100 and 65. The second test peaked at 74 percent with no timeouts.

More agents for database administrators

Quality control · IT and development

Runbook Freshness Test Agent

Runbooks that are tested, current and dated

10 steps1 check7 jobs
Decisions · IT and development

Index Usage and Redundancy Review Agent

Fewer needless indexes with every drop tested against real queries

11 steps2 checks3 jobs
Analysis · IT and development

Replication Lag Investigation Agent

A confirmed cause of replica lag and a tested fix, with the evidence recorded

10 steps2 checks2 jobs
Risk management · Quality control

Patch and Upgrade Rehearsal Agent

A patch that is proven on a realistic copy, with known issues and timings before production

10 steps2 checks1 job