Inhaltsverzeichnis

Runbook: Health Check

Runbook: Health Check

Trajanje: ~2 minute
Uloga: Gateway-Operator
Ucestalost: Dnevno / Automatizirano

Provjera dostupnosti i funkcionalnosti Gatewaya.

Tijek rada

flowchart TD A[Health Check pokrenuti] --> B[/health Endpoint] B --> C{Healthy?} C -->|Da| D[API test] C -->|Ne| E[Logove provjeriti] D --> F{Podaci vraceni?} F -->|Da| G[OK - Gotovo] F -->|Ne| H[DSN provjeriti] E --> I[Server ponovno pokrenuti] H --> I style G fill:#e8f5e9 style E fill:#ffebee style H fill:#fff3e0

1. Basic Health Check

# Jednostavan Health Check
curl -s -o /dev/null -w "%{http_code}" http://localhost:5000/health
 
# Ocekivani odgovor: 200

# S Response-Body
curl -s http://localhost:5000/health
 
# Ocekivani odgovor: "Healthy"

2. Extended Health Check

# Swagger dostupan?
curl -s -o /dev/null -w "%{http_code}" http://localhost:5000/swagger
 
# API verziju provjeriti
curl -s http://localhost:5000/api/v1/info | jq

3. DSN konnektivnost

# Sve DSN proci
for dsn in demo produktion reporting; do
  echo "Testing $dsn..."
  curl -s -o /dev/null -w "$dsn: %{http_code}\n" \
    "http://localhost:5000/api/v1/dsn/$dsn/tables"
done

PowerShell verzija:

$dsns = @("demo", "produktion", "reporting")
foreach ($dsn in $dsns) {
    $result = Invoke-WebRequest -Uri "http://localhost:5000/api/v1/dsn/$dsn/tables" -UseBasicParsing
    Write-Host "$dsn : $($result.StatusCode)"
}

4. Vrijeme odgovora mjeriti

# Pojedinacni upit
curl -s -o /dev/null -w "Vrijeme: %{time_total}s\n" \
  "http://localhost:5000/api/v1/dsn/demo/tables/Products?\$top=10"
 
# Vise prolaza
for i in {1..5}; do
  curl -s -o /dev/null -w "%{time_total}\n" \
    "http://localhost:5000/api/v1/dsn/demo/tables/Products?\$top=10"
done | awk '{sum+=$1} END {print "Prosjek: " sum/NR "s"}'

5. Kontrolna lista

#	Provjera	Ocekivanje	Da/Ne
—	———–	———–	—
1	/health	200 + „Healthy“	-
2	/swagger	200	-
3	DSN „demo“ dostupna	200	-
4	Vrijeme odgovora	< 1s	-
5	Nema gresaka u logovima	Nema ERROR	-

Automatizirani Health Check

Cron (Linux):

# /etc/cron.d/gateway-health
*/5 * * * * root curl -sf http://localhost:5000/health || systemctl restart data-gateway

Scheduled Task (Windows):

# health-check.ps1
$response = Invoke-WebRequest -Uri "http://localhost:5000/health" -UseBasicParsing -TimeoutSec 5
if ($response.StatusCode -ne 200) {
    Restart-Service -Name "DataGateway" -Force
    Send-MailMessage -To "admin@example.com" -Subject "Gateway Restart" -Body "Gateway je automatski ponovno pokrenut"
}

Rjesavanje problema

Problem	Uzrok	Rjesenje
———	———	——–
`Connection refused`	Gateway nije pokrenut	Server pokrenuti
`503 Service Unavailable`	Startup jos nije zavrsen	30s pricekati, ponovno provjeriti
`500 Internal Server Error`	Config greska	Logove provjeriti
Timeout	Gateway preopterecen	Opterecenje smanjiti, resurse povecati

Pragovi

Metrika	Zeleno	Zuto	Crveno
——–	——	——	—–
Vrijeme odgovora	< 500ms	500ms-2s	> 2s
Error-Rate	< 1%	1-5%	> 5%
CPU	< 50%	50-80%	> 80%
Memory	< 70%	70-90%	> 90%

Povezani runbookovi

Server pokrenuti - Kod ispada
Provjera logova - Analiza gresaka
Prometheus - Automatski monitoring

« <- DSN upravljanje | -> Provjera logova »

Wolfgang van der Stille @ EMSR DATA d.o.o. - Data Gateway Professional

operator, runbook, health, check, monitoring