AI LLM Creation and Usage

Overview

What follows are the essential steps to create an AI LLM in Redhat\OEL\Alma 9 Linux.

Ensure CA Set

-- Install
   sudo dnf install -y \
   ca-certificates \
   p11-kit \
   p11-kit-trust \
   openssl

  sudo update-ca-trust



-- QC
ls -l /etc/pki/tls/certs/ca-bundle.crt
curl -I https://www.google.com
openssl s_client -connect ollama.com:443 -showcerts </dev/null
  Verify return code: 0 (ok)

Force-refresh Mozilla trust bundle

 sudo dnf reinstall -y ca-certificates
 sudo update-ca-trust extract

DL Ollama

 https://ollama.com/download/ollama-linux-amd64.tar.zst
 cp /media/sf_sw/Tools/ollama-linux-amd64.tar.zst /u01/sw/

Extract

cd /u01/sw
tar -xvf ollama-linux-amd64.tar.zst bin/ollama
mv bin/ollama .
rmdir bin
ls -lh ollama

Install

 mv ollama /usr/local/bin/
 chmod +x /usr/local/bin/ollama

Start Server (from a diff console)

ollama serve
ollama --version
ex: llama version is 0.18.2

-- 101 Tests
ss -lntp | grep 11434

Pull a Model

   ---- Largest
   ollama pull gemma3
   ollama list


   ---- Adequate
   ollama pull llama3
   ollama list

Make Model Active

--- For gemma3
curl -s http://127.0.0.1:11434/api/generate \
  -H 'Content-Type: application/json' \
  -d '{
    "model": "gemma3",
    "prompt": "Say hello in one short sentence."
  }'


---- For llama3
curl http://localhost:11434/api/generate -d '{
  "model": "llama3",
  "prompt": "Summarize this contract clause..."
}'


-- Manual Test Example
curl -s http://127.0.0.1:11434/api/chat \
  -H 'Content-Type: application/json' \
  -d '{
    "model": "llama3",
    "messages": [
      {"role":"system","content":"Be concise."},
      {"role":"user","content":"What is AlmaLinux?"}
    ]
  }'

Post Reboot Essential Commands to Use LLM

1) Verify Ollama is running
   systemctl status ollama --no-pager -l
   ✅ Active: active (running)


2) Confirm API is up
   curl -s http://127.0.0.1:11434/api/version
   ✅ Should return JSON


3) Check models are still there
   ollama list


4) Use it (two main ways)
   🟢 Interactive (fastest sanity use)
      ollama run llama3

   🟢 API (what you’ll actually use in apps)
      curl -s http://127.0.0.1:11434/api/chat \
        -H 'Content-Type: application/json' \
        -d '{
          "model": "llama3",
          "messages": [
            {"role":"system","content":"Be concise."},
            {"role":"user","content":"What is AlmaLinux?"}
          ]
        }' | jq

✅ If something doesn’t work after reboot

 systemctl status ollama --no-pager -l
 journalctl -u ollama -n 50 --no-pager
 ss -lntp | grep 11434

✅ Optional (recommended for production feel)

curl -s http://127.0.0.1:11434/api/generate -d '{
  "model": "llama3",
  "prompt": "ping"
}' > /dev/null