Loading...

XML

Word

Printable

Type: Bug
Resolution: Done
Priority: Undefined
Fix Version/s: v0.2.1
Affects Version/s: v0.1.8
Component/s: insights-on-prem
Labels:
- cost-onprem-0.2
- qe

Blocked:
False
Blocked Reason:

Hide

None

Show
None
Ready:
False
Intelligence Requested:
Market:

SFDC Cases Counter:
SFDC Cases Open:
SFDC Cases Links:

Summary of Issues

ROS API Pod Label Mismatch (install-helm-chart.sh line 841)
Kruize Service Name Mismatch (install-helm-chart.sh line 892)
Service URL Construction - Missing ros-ocp infix (test-ocp-dataflow-jwt.sh lines 457-497)
Database Pod Label Selector - Wrong label (test-ocp-dataflow-jwt.sh line 746, query-kruize.sh line 50)
Database Credentials Secret Name - Missing ros-ocp infix (test-ocp-dataflow-jwt.sh line 755, query-kruize.sh line 12)
Database Username Extraction - Trying to extract hardcoded value from secret (test-ocp-dataflow-jwt.sh lines 756-766, query-kruize.sh line 13)

Description

Multiple scripts in the cost-onprem-chart repository use incorrect resource naming patterns, causing both false-positive health check failures and complete test failures. These issues appear to be related to incomplete updates following the chart rename from "ros-helm-chart" to "cost-onprem-chart".

Impact

Severity: High - Test scripts completely fail due to incorrect service names and database access
User Experience:
- Health checks report false failures but deployment actually succeeds
- End-to-end test scripts fail with HTTP 000 errors (cannot reach services)
- Database verification queries fail (cannot find database pod or credentials)
- Confusing error messages during installation/testing
Affected Scripts:
- scripts/install-helm-chart.sh - Health check functions
- scripts/test-ocp-dataflow-jwt.sh - End-to-end dataflow testing and database verification
- scripts/query-kruize.sh - Kruize database query utility
Installation/Testing Outcome: Deployments succeed but health checks fail; test scripts completely fail

Root Cause

The scripts use outdated naming patterns that don't match the actual Helm chart resource names after the chart was renamed. The Helm fullname template now generates names with ros-ocp infix, but scripts weren't updated.

Affected Scripts and Issues

install-helm-chart.sh Issues

Issue 1: ROS API Pod Label Mismatch (Line 841)

Script searches for: app.kubernetes.io/name=ros-api
Actual pod label: app.kubernetes.io/name=rosocp-api
Result: "ROS API pod not found" error

Issue 2: Kruize Service Name Mismatch (Line 892)

Script searches for: svc/cost-onprem-kruize
Actual service name: svc/cost-onprem-ros-ocp-kruize
Result: "Kruize API service is not responding" error

test-ocp-dataflow-jwt.sh Issues

Service URL Construction Issues

Issue 3: Service URL Construction (get_service_url() function, lines 457-497)

Function constructs: $HELM_RELEASE_NAME-$service_name
Should construct: $HELM_RELEASE_NAME-ros-ocp-$service_name
Examples:
- Tries to find: cost-onprem-ingress
- Actually exists: cost-onprem-ros-ocp-ingress
- Tries to find: cost-onprem-rosocp-api
- Actually exists: cost-onprem-ros-ocp-rosocp-api
Result: HTTP 000 errors (cannot connect), complete test failure

Database Access Issues (Discovered during dataflow test)

Issue 4: Database Pod Label Selector (Line 746)

Script searches for: app.kubernetes.io/name=database
Actual pod label: app.kubernetes.io/name=db-kruize
Result: "Database pod not found" error

Issue 5: Database Credentials Secret Name (Line 755)

Script searches for: cost-onprem-db-credentials
Actual secret name: cost-onprem-ros-ocp-db-credentials
Result: "Unable to retrieve Kruize database credentials" error

Issue 6: Database Username Extraction (Lines 756-766)

Script tries to extract: kruize-user from secret
Reality: Secret only contains passwords (kruize-password, ros-password, sources-password), username is hardcoded in Helm chart as "postgres"
Result: Username variable is empty, database connection fails

query-kruize.sh Issues

The same database access issues (Issues 4, 5, 6) exist in query-kruize.sh:

Issue 4: Database Pod Label Selector (Line 50)

Script searches for: app.kubernetes.io/name=database
Actual pod label: app.kubernetes.io/name=db-kruize
Result: "Kruize database pod not found" error

Issue 5: Database Credentials Secret Name (Line 12)

Script searches for: cost-onprem-db-credentials
Actual secret name: cost-onprem-ros-ocp-db-credentials
Result: "Unable to retrieve Kruize database credentials" error

Issue 6: Database Username Extraction (Line 13)

Script tries to extract: kruize-user from secret
Reality: Secret only contains passwords, username is hardcoded to "postgres"
Result: Script fails with error about missing credentials

Evidence

Health Check False Positives (install-helm-chart.sh)

[INFO] Running health checks...
[INFO] Testing internal service connectivity...
[ERROR] ✗ ROS API pod not found
[INFO] Testing services via port-forwarding (OpenShift approach)...
[INFO] Testing Ingress API via port-forward...
[SUCCESS] ✓ Ingress API service is healthy (port-forward, internal port 8081)
[INFO] Testing Kruize API via port-forward...
[ERROR] ✗ Kruize API service is not responding (port-forward)
[INFO] Testing external route accessibility (informational)...
[SUCCESS]   → ROS API externally accessible: http://cost-onprem-ros-ocp-main-cost-onprem.apps.ocp-test.qe.lab.redhat.com/status
[SUCCESS]   → 1 route(s) externally accessible
[ERROR] 2 core service check(s) failed

Test Script Complete Failure (test-ocp-dataflow-jwt.sh)

[INFO] === Preflight: JWT Authentication Validation ===
[INFO] Testing ingress at: http://cost-onprem-ingress.cost-onprem.svc.cluster.local:8080
[INFO] Test 1: Request without JWT token
[WARNING]   ⚠ Expected 401, got 000000 (may indicate route/service issue)

[INFO] === STEP 1: Upload Test Data with JWT Authentication ====
[INFO] Uploading to: http://cost-onprem-ingress.cost-onprem.svc.cluster.local:8080/v1/upload
[ERROR] Upload failed with HTTP 000
[ERROR] Response:
[ERROR] Upload with JWT authentication failed

Database Access Failure

[ERROR] Database pod not found

[ERROR] Unable to retrieve Kruize database credentials from secret 'cost-onprem-db-credentials'

Actual Service Names (What Really Exists)

$ oc get svc -n cost-onprem | grep -E "api|ingress|kruize"
cost-onprem-ros-ocp-db-kruize                      ClusterIP      172.30.14.216    [none]        5432/TCP                     3h10m
cost-onprem-ros-ocp-ingress                        ClusterIP      172.30.82.140    [none]        8080/TCP,9901/TCP            3h10m
cost-onprem-ros-ocp-kruize                         ClusterIP      172.30.53.128    [none]        8080/TCP                     3h10m
cost-onprem-ros-ocp-rosocp-api                     ClusterIP      172.30.105.65    [none]        8000/TCP,9901/TCP,9000/TCP   3h10m

Actual Route Names

$ oc get routes -n cost-onprem
NAME                          HOST/PORT
cost-onprem-ros-ocp-ingress   cost-onprem-ros-ocp-ingress-cost-onprem.apps.ocp-test.qe.lab.redhat.com
cost-onprem-ros-ocp-main      cost-onprem-ros-ocp-main-cost-onprem.apps.ocp-test.qe.lab.redhat.com

Actual Pod Status (All Healthy)

$ oc get pods -n cost-onprem
NAME                                                              READY   STATUS    RESTARTS   AGE
cost-onprem-ros-ocp-rosocp-api-758c6c65c8-lk482                   2/2     Running   0          3h12m
cost-onprem-ros-ocp-ingress-889b69c5c-59697                       2/2     Running   0          3h12m
cost-onprem-ros-ocp-kruize-64644796fd-v7hck                       1/1     Running   0          3h12m
[... all 12 pods running successfully ...]

Actual Pod Labels

ROS API Pod:

$ oc get pods -n cost-onprem cost-onprem-ros-ocp-rosocp-api-758c6c65c8-lk482 -o jsonpath='{.metadata.labels.app\.kubernetes\.io/name}'
rosocp-api

Database Pod:

$ oc get pods -n cost-onprem -l app.kubernetes.io/name=db-kruize -o jsonpath='{.items[0].metadata.name}'
cost-onprem-ros-ocp-db-kruize-0

Actual Database Secret

$ oc get secret cost-onprem-ros-ocp-db-credentials -n cost-onprem -o jsonpath='{.data}'
{"kruize-password":"...","ros-password":"...","sources-password":"..."}

Note: Secret contains only passwords, not usernames (usernames are hardcoded in Helm chart).

Verification that Services Actually Work

# ROS API works internally
$ oc exec -n cost-onprem cost-onprem-ros-ocp-rosocp-api-758c6c65c8-lk482 -c rosocp-api -- curl -s http://localhost:8000/status
{"api-server":"working"}

# Kruize works internally
$ oc exec -n cost-onprem cost-onprem-ros-ocp-kruize-64644796fd-v7hck -- curl -s http://localhost:8080/listPerformanceProfiles | head -5
[
  {
    "name": "resource-optimization-openshift",
    "profile_version": 1.0,
    "k8s_type": "openshift",

Steps to Reproduce

Health Check Issue

Deploy on OpenShift cluster with JWT authentication enabled
Run: export JWT_AUTH_ENABLED=true && ./install-helm-chart.sh
Wait for installation to complete
Observe health check output shows false errors despite successful deployment

Test Script Issue

Complete successful deployment as above
Create service account with admin access
Login as service account: oc login --token=... --server=...
Run: ./scripts/test-ocp-dataflow-jwt.sh
Observe HTTP 000 errors when trying to connect to services
Observe database access errors when trying to verify Kruize recommendations
Verify actual services exist with correct ros-ocp naming

Current Buggy Code

install-helm-chart.sh

Line 841 - Wrong Pod Label:

local api_pod=$(kubectl get pods -n "$NAMESPACE" -l app.kubernetes.io/name=ros-api -o jsonpath='{.items[0].metadata.name}' 2>/dev/null)

Line 892 - Wrong Service Name:

kubectl port-forward -n "$NAMESPACE" svc/cost-onprem-kruize 18081:8080 --request-timeout=90s >/dev/null 2>&1 &

test-ocp-dataflow-jwt.sh

Lines 462 & 494 - Wrong Service/Route Name Construction:

get_service_url() {
    local service_name="$1"
    local path="$2"

    # Try to get OpenShift route first
    local route_name="$HELM_RELEASE_NAME-$service_name"  # WRONG

    # ... route lookup ...

    # Fallback to service (for port-forward or internal access)
    local service_port=$(oc get svc "$HELM_RELEASE_NAME-$service_name" -n "$NAMESPACE" -o jsonpath='{.spec.ports[0].port}' 2>/dev/null || echo "8080")  # WRONG
    echo "http://$HELM_RELEASE_NAME-$service_name.$NAMESPACE.svc.cluster.local:${service_port}$path"  # WRONG
}

Line 746 - Wrong Database Pod Label:

local db_pod=$(oc get pods -n "$NAMESPACE" -l "app.kubernetes.io/name=database" -o jsonpath='{.items[0].metadata.name}' 2>/dev/null)

Line 755 - Wrong Database Credentials Secret Name:

local db_secret_name="cost-onprem-db-credentials"

Lines 756-757 - Wrong Database Username Extraction:

local kruize_user=$(oc get secret -n "$NAMESPACE" "$db_secret_name" -o jsonpath='{.data.kruize-user}' 2>/dev/null | base64 -d)
local kruize_password=$(oc get secret -n "$NAMESPACE" "$db_secret_name" -o jsonpath='{.data.kruize-password}' 2>/dev/null | base64 -d)

query-kruize.sh

Line 50 - Wrong Database Pod Label:

local db_pod=$(oc get pods -n "$NAMESPACE" -l "app.kubernetes.io/name=database" -o jsonpath='{.items[0].metadata.name}' 2>/dev/null)

Line 12 - Wrong Database Credentials Secret Name:

DB_SECRET_NAME="${DB_SECRET_NAME:-cost-onprem-db-credentials}"

Lines 13-14 - Wrong Database Username Extraction:

KRUIZE_USER=$(oc get secret -n "$NAMESPACE" "$DB_SECRET_NAME" -o jsonpath='{.data.kruize-user}' 2>/dev/null | base64 -d)
KRUIZE_PASSWORD=$(oc get secret -n "$NAMESPACE" "$DB_SECRET_NAME" -o jsonpath='{.data.kruize-password}' 2>/dev/null | base64 -d)

Expected/Fixed Code

install-helm-chart.sh

Line 841 - Correct Pod Label:

local api_pod=$(kubectl get pods -n "$NAMESPACE" -l app.kubernetes.io/name=rosocp-api -o jsonpath='{.items[0].metadata.name}' 2>/dev/null)

Line 892 - Correct Service Name:

kubectl port-forward -n "$NAMESPACE" svc/cost-onprem-ros-ocp-kruize 18081:8080 --request-timeout=90s >/dev/null 2>&1 &

test-ocp-dataflow-jwt.sh

Lines 462 & 494 - Correct Service/Route Name Construction:

get_service_url() {
    local service_name="$1"
    local path="$2"

    # Try to get OpenShift route first
    local route_name="$HELM_RELEASE_NAME-ros-ocp-$service_name"  # FIXED

    # ... route lookup ...

    # Fallback to service (for port-forward or internal access)
    local service_port=$(oc get svc "$HELM_RELEASE_NAME-ros-ocp-$service_name" -n "$NAMESPACE" -o jsonpath='{.spec.ports[0].port}' 2>/dev/null || echo "8080")  # FIXED
    echo "http://$HELM_RELEASE_NAME-ros-ocp-$service_name.$NAMESPACE.svc.cluster.local:${service_port}$path"  # FIXED
}

Note: The special case handling for "ros-api" → "main" route (lines 464-476) also needs similar updates to use ros-ocp prefix.

Line 746 - Correct Database Pod Label:

local db_pod=$(oc get pods -n "$NAMESPACE" -l "app.kubernetes.io/name=db-kruize" -o jsonpath='{.items[0].metadata.name}' 2>/dev/null)

Line 755 - Correct Database Credentials Secret Name:

local db_secret_name="${HELM_RELEASE_NAME}-ros-ocp-db-credentials"

Lines 756-766 - Correct Database Credentials Extraction:

# Extract Kruize database credentials from secret
local db_secret_name="${HELM_RELEASE_NAME}-ros-ocp-db-credentials"
local kruize_password=$(oc get secret -n "$NAMESPACE" "$db_secret_name" -o jsonpath='{.data.kruize-password}' 2>/dev/null | base64 -d)

# Username is hardcoded to "postgres" in the Helm chart
local kruize_user="postgres"
local kruize_db="kruize_db"

if [ -z "$kruize_password" ]; then
    echo_error "Unable to retrieve Kruize database password from secret '$db_secret_name'"
    echo_info "Use './query-kruize.sh --cluster $cluster_id' to check recommendations later"
    return 1
fi

query-kruize.sh

Line 50 - Correct Database Pod Label:

local db_pod=$(oc get pods -n "$NAMESPACE" -l "app.kubernetes.io/name=db-kruize" -o jsonpath='{.items[0].metadata.name}' 2>/dev/null)

Line 12 - Correct Database Credentials Secret Name:

DB_SECRET_NAME="${DB_SECRET_NAME:-cost-onprem-ros-ocp-db-credentials}"

Lines 13-22 - Correct Database Credentials Extraction:

# Username is hardcoded to "postgres" in the Helm chart
KRUIZE_USER="postgres"
KRUIZE_PASSWORD=$(oc get secret -n "$NAMESPACE" "$DB_SECRET_NAME" -o jsonpath='{.data.kruize-password}' 2>/dev/null | base64 -d)
KRUIZE_DB="kruize_db"

if [ -z "$KRUIZE_PASSWORD" ]; then
    echo "Error: Unable to retrieve Kruize database password from secret '$DB_SECRET_NAME'" >&2
    echo "Ensure the secret exists and contains 'kruize-password' key" >&2
    exit 1
fi

Workaround

install-helm-chart.sh

The health check failures are informational only and can be ignored. All services are actually working correctly despite the error messages. Users can verify deployment success by:

# Check all pods are running
oc get pods -n cost-onprem

# Test ROS API
oc exec -n cost-onprem -l app.kubernetes.io/name=rosocp-api -c rosocp-api -- curl -s http://localhost:8000/status

# Test Kruize API
oc exec -n cost-onprem -l app.kubernetes.io/name=kruize -- curl -s http://localhost:8080/listPerformanceProfiles

test-ocp-dataflow-jwt.sh

Manually test the services using the correct names:

# Test ingress route directly
curl -k https://cost-onprem-ros-ocp-ingress-cost-onprem.apps.ocp-test.qe.lab.redhat.com/ready

# Or use port-forward to test upload
oc port-forward -n cost-onprem svc/cost-onprem-ros-ocp-ingress 8080:8080 &
curl -X POST -F "file=@test.tar.gz" -H "Authorization: Bearer $JWT_TOKEN" http://localhost:8080/v1/upload

# Verify Kruize database manually
oc exec -n cost-onprem cost-onprem-ros-ocp-db-kruize-0 -- \
  psql -U postgres -d kruize_db -c "SELECT COUNT(*) FROM kruize_experiments;"

History

Previous Partial Fix (Incomplete):

Commit 7ca133e (Nov 10, 2025): "Fix wrong ros-api service name"
Fixed typo: ross-api → ros-api
Still Incorrect: Should have been changed to rosocp-api

This suggests the naming pattern was updated incompletely during the chart rename work.

Related Issues

~~FLPATH-2891~~: Storage credentials secret naming regression (same root cause - chart rename)
Commit 291b59f: "Rename chart name to cost-onprem"
Commit 187108f: "Use cost-onprem as chart name and default namespace"

All bugs (~~FLPATH-2891~~, ~~FLPATH-2892~~) stem from incomplete resource name updates following the chart rename from "ros-helm-chart" to "cost-onprem-chart".

Environment Details

Repository: https://github.com/insights-onprem/cost-onprem-chart
Git Commit: e5d6a2d1b82d0fcaf8594cfae217f61209c10cc2
Chart Version: v0.1.8-41-ge5d6a2d
OpenShift Version: 4.18.26
Kubernetes Version: v1.31.13
Helm Release Name: cost-onprem (default)
Namespace: cost-onprem (default)
Platform: OpenShift
JWT Auth: Enabled
Affected Scripts:
- scripts/install-helm-chart.sh
- scripts/test-ocp-dataflow-jwt.sh
- scripts/query-kruize.sh

Proposed Fix

Update all three scripts to use the correct ros-ocp naming convention:

install-helm-chart.sh

Unable to find source-code formatter for language: diff. Available languages are: actionscript, ada, applescript, bash, c, c#, c++, cpp, css, erlang, go, groovy, haskell, html, java, javascript, js, json, lua, none, nyan, objc, perl, php, python, r, rainbow, ruby, scala, sh, sql, swift, visualbasic, xml, yaml

# Line 841: Change pod label search
-        local api_pod=$(kubectl get pods -n "$NAMESPACE" -l app.kubernetes.io/name=ros-api -o jsonpath='{.items[0].metadata.name}' 2>/dev/null)
+        local api_pod=$(kubectl get pods -n "$NAMESPACE" -l app.kubernetes.io/name=rosocp-api -o jsonpath='{.items[0].metadata.name}' 2>/dev/null)

# Line 892: Change service name
-        kubectl port-forward -n "$NAMESPACE" svc/cost-onprem-kruize 18081:8080 --request-timeout=90s >/dev/null 2>&1 &
+        kubectl port-forward -n "$NAMESPACE" svc/cost-onprem-ros-ocp-kruize 18081:8080 --request-timeout=90s >/dev/null 2>&1 &

test-ocp-dataflow-jwt.sh

# Update get_service_url() function to add -ros-ocp- infix

# Line 462: Update route name construction
-    local route_name="$HELM_RELEASE_NAME-$service_name"
+    local route_name="$HELM_RELEASE_NAME-ros-ocp-$service_name"

# Line 467: Update "main" route special case
-        route_name="$HELM_RELEASE_NAME-main"
+        route_name="$HELM_RELEASE_NAME-ros-ocp-main"

# Line 474: Update ros-api route fallback
-        route_name="$HELM_RELEASE_NAME-ros-api"
+        route_name="$HELM_RELEASE_NAME-ros-ocp-rosocp-api"

# Line 494: Update service name construction
-        local service_port=$(oc get svc "$HELM_RELEASE_NAME-$service_name" -n "$NAMESPACE" -o jsonpath='{.spec.ports[0].port}' 2>/dev/null || echo "8080")
+        local service_port=$(oc get svc "$HELM_RELEASE_NAME-ros-ocp-$service_name" -n "$NAMESPACE" -o jsonpath='{.spec.ports[0].port}' 2>/dev/null || echo "8080")

# Line 495: Update service URL construction
-        echo "http://$HELM_RELEASE_NAME-$service_name.$NAMESPACE.svc.cluster.local:${service_port}$path"
+        echo "http://$HELM_RELEASE_NAME-ros-ocp-$service_name.$NAMESPACE.svc.cluster.local:${service_port}$path"

# Line 746: Fix database pod label
-local db_pod=$(oc get pods -n "$NAMESPACE" -l "app.kubernetes.io/name=database" -o jsonpath='{.items[0].metadata.name}' 2>/dev/null)
+local db_pod=$(oc get pods -n "$NAMESPACE" -l "app.kubernetes.io/name=db-kruize" -o jsonpath='{.items[0].metadata.name}' 2>/dev/null)

# Line 755: Fix database credentials secret name
-local db_secret_name="cost-onprem-db-credentials"
+local db_secret_name="${HELM_RELEASE_NAME}-ros-ocp-db-credentials"

# Lines 756-766: Fix database credentials extraction
-local kruize_user=$(oc get secret -n "$NAMESPACE" "$db_secret_name" -o jsonpath='{.data.kruize-user}' 2>/dev/null | base64 -d)
-local kruize_password=$(oc get secret -n "$NAMESPACE" "$db_secret_name" -o jsonpath='{.data.kruize-password}' 2>/dev/null | base64 -d)
+# Extract Kruize database credentials from secret
+local db_secret_name="${HELM_RELEASE_NAME}-ros-ocp-db-credentials"
+local kruize_password=$(oc get secret -n "$NAMESPACE" "$db_secret_name" -o jsonpath='{.data.kruize-password}' 2>/dev/null | base64 -d)
+
+# Username is hardcoded to "postgres" in the Helm chart
+local kruize_user="postgres"
+local kruize_db="kruize_db"
+
+if [ -z "$kruize_password" ]; then
+    echo_error "Unable to retrieve Kruize database password from secret '$db_secret_name'"
+    echo_info "Use './query-kruize.sh --cluster $cluster_id' to check recommendations later"
+    return 1
+fi

query-kruize.sh

# Line 50: Fix database pod label
-    local db_pod=$(oc get pods -n "$NAMESPACE" -l "app.kubernetes.io/name=database" -o jsonpath='{.items[0].metadata.name}' 2>/dev/null)
+    local db_pod=$(oc get pods -n "$NAMESPACE" -l "app.kubernetes.io/name=db-kruize" -o jsonpath='{.items[0].metadata.name}' 2>/dev/null)

# Line 12: Fix database credentials secret name
-DB_SECRET_NAME="${DB_SECRET_NAME:-cost-onprem-db-credentials}"
+DB_SECRET_NAME="${DB_SECRET_NAME:-cost-onprem-ros-ocp-db-credentials}"

# Lines 13-22: Fix database credentials extraction
-KRUIZE_USER=$(oc get secret -n "$NAMESPACE" "$DB_SECRET_NAME" -o jsonpath='{.data.kruize-user}' 2>/dev/null | base64 -d)
-KRUIZE_PASSWORD=$(oc get secret -n "$NAMESPACE" "$DB_SECRET_NAME" -o jsonpath='{.data.kruize-password}' 2>/dev/null | base64 -d)
-KRUIZE_DB="kruize_db"
-
-if [ -z "$KRUIZE_USER" ] || [ -z "$KRUIZE_PASSWORD" ]; then
-    echo "Error: Unable to retrieve Kruize database credentials from secret '$DB_SECRET_NAME'" >&2
-    echo "Ensure the secret exists and contains 'kruize-user' and 'kruize-password' keys" >&2
-    exit 1
-fi
+# Username is hardcoded to "postgres" in the Helm chart
+KRUIZE_USER="postgres"
+KRUIZE_PASSWORD=$(oc get secret -n "$NAMESPACE" "$DB_SECRET_NAME" -o jsonpath='{.data.kruize-password}' 2>/dev/null | base64 -d)
+KRUIZE_DB="kruize_db"
+
+if [ -z "$KRUIZE_PASSWORD" ]; then
+    echo "Error: Unable to retrieve Kruize database password from secret '$DB_SECRET_NAME'" >&2
+    echo "Ensure the secret exists and contains 'kruize-password' key" >&2
+    exit 1
+fi

These changes align all three scripts with the actual Helm chart resource naming conventions established by the fullname template.

Assignee:: Moti Asayag

Reporter:: Chad Crum

Votes:: 0 Vote for this issue

Watchers:: 1 Start watching this issue

Created:: 2025/11/19 6:18 PM

Updated:: 2025/12/01 4:38 PM

Resolved:: 2025/12/01 4:38 PM

Details

Description

Summary of Issues

Description

Impact

Root Cause

Affected Scripts and Issues

install-helm-chart.sh Issues

test-ocp-dataflow-jwt.sh Issues

Service URL Construction Issues

Database Access Issues (Discovered during dataflow test)

query-kruize.sh Issues

Evidence

Health Check False Positives (install-helm-chart.sh)

Test Script Complete Failure (test-ocp-dataflow-jwt.sh)

Database Access Failure

Actual Service Names (What Really Exists)

Actual Route Names

Actual Pod Status (All Healthy)

Actual Pod Labels

Actual Database Secret

Verification that Services Actually Work

Steps to Reproduce

Health Check Issue

Test Script Issue

Current Buggy Code

install-helm-chart.sh

test-ocp-dataflow-jwt.sh

query-kruize.sh

Expected/Fixed Code

install-helm-chart.sh

test-ocp-dataflow-jwt.sh

query-kruize.sh

Workaround

install-helm-chart.sh

test-ocp-dataflow-jwt.sh

History

Related Issues

Environment Details

Proposed Fix

install-helm-chart.sh

test-ocp-dataflow-jwt.sh

query-kruize.sh

Attachments

Easy Agile Planning Poker

Activity

People

Dates