From 9a64644670afaa9ca23fe54d774056db5c8c1927 Mon Sep 17 00:00:00 2001
From: =?UTF-8?q?Jan=20G=C3=B3recki?= <jan.gorecki@shopsys.com>
Date: Wed, 19 Aug 2026 02:13:24 +0200
Subject: [PATCH 1/5] Drop the legacy sleep 30 and bound the hook Jobs

The first-deploy migration command no longer sleeps 30 seconds - the
shopsys-infra release is installed with wait: true before the hook
runs, so readiness is guaranteed. The migration and post-deploy Jobs
get configurable activeDeadlineSeconds (1800/600 s defaults) so a
stuck Job fails cleanly instead of blocking the deploy until the
global helmfile timeout.

Closes #11

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
---
 .../templates/hooks/job-migrate-application.yaml   |  7 +++++--
 .../templates/hooks/job-post-deploy.yaml           |  1 +
 charts/shopsys-app/tests/hooks_test.yaml           |  8 +++++++-
 charts/shopsys-app/values.schema.json              | 14 ++++++++++++++
 charts/shopsys-app/values.yaml                     |  3 +++
 docs/migrating-from-shopsys-deployment.md          |  7 +++++--
 .../basic-production/expected/continuous.yaml      |  4 ++++
 .../expected/first-deploy-with-demo-data.yaml      |  6 +++++-
 .../basic-production/expected/first-deploy.yaml    |  6 +++++-
 .../expected/continuous.yaml                       |  4 ++++
 .../expected/first-deploy-with-demo-data.yaml      |  6 +++++-
 .../expected/first-deploy.yaml                     |  6 +++++-
 .../expected/continuous.yaml                       |  4 ++++
 .../expected/first-deploy-with-demo-data.yaml      |  6 +++++-
 .../expected/first-deploy.yaml                     |  6 +++++-
 .../escaping-env/expected/continuous.yaml          |  4 ++++
 .../expected/first-deploy-with-demo-data.yaml      |  6 +++++-
 .../escaping-env/expected/first-deploy.yaml        |  6 +++++-
 .../expected/continuous.yaml                       |  4 ++++
 .../expected/first-deploy-with-demo-data.yaml      |  6 +++++-
 .../expected/first-deploy.yaml                     |  6 +++++-
 21 files changed, 105 insertions(+), 15 deletions(-)

diff --git a/charts/shopsys-app/templates/hooks/job-migrate-application.yaml b/charts/shopsys-app/templates/hooks/job-migrate-application.yaml
index 08db7d3..03c2cf5 100644
--- a/charts/shopsys-app/templates/hooks/job-migrate-application.yaml
+++ b/charts/shopsys-app/templates/hooks/job-migrate-application.yaml
@@ -18,9 +18,9 @@ The command variant is selected by deploy.firstDeploy.*:
 {{- $command := printf "cd /var/www/html && ./phing %s" $targets.continuous }}
 {{- if .Values.deploy.firstDeploy.enabled }}
 {{- if .Values.deploy.firstDeploy.loadDemoData }}
-{{- $command = printf "cd /var/www/html && sleep 30 && ./phing %s" $targets.firstDeployWithDemoData }}
+{{- $command = printf "cd /var/www/html && ./phing %s" $targets.firstDeployWithDemoData }}
 {{- else }}
-{{- $command = printf "cd /var/www/html && sleep 30 && ./phing %s" $targets.firstDeploy }}
+{{- $command = printf "cd /var/www/html && ./phing %s" $targets.firstDeploy }}
 {{- end }}
 {{- end }}
 apiVersion: batch/v1
@@ -35,6 +35,9 @@ metadata:
     helm.sh/hook-delete-policy: before-hook-creation
 spec:
   backoffLimit: 0
+  # Bound the migration - a stuck Job otherwise blocks the deploy until the global
+  # helmfile timeout and stays running in the cluster
+  activeDeadlineSeconds: {{ .Values.deploy.migration.activeDeadlineSeconds }}
   template:
     spec:
       volumes:
diff --git a/charts/shopsys-app/templates/hooks/job-post-deploy.yaml b/charts/shopsys-app/templates/hooks/job-post-deploy.yaml
index bfa5c40..55c4505 100644
--- a/charts/shopsys-app/templates/hooks/job-post-deploy.yaml
+++ b/charts/shopsys-app/templates/hooks/job-post-deploy.yaml
@@ -17,6 +17,7 @@ metadata:
     helm.sh/hook-delete-policy: before-hook-creation
 spec:
   backoffLimit: 0
+  activeDeadlineSeconds: {{ .Values.deploy.postDeploy.activeDeadlineSeconds }}
   template:
     spec:
       volumes:
diff --git a/charts/shopsys-app/tests/hooks_test.yaml b/charts/shopsys-app/tests/hooks_test.yaml
index 6ce6906..9867f2c 100644
--- a/charts/shopsys-app/tests/hooks_test.yaml
+++ b/charts/shopsys-app/tests/hooks_test.yaml
@@ -25,6 +25,9 @@ tests:
       - matchRegex:
           path: spec.template.spec.containers[0].command[2]
           pattern: 'db-migrations-count-with-maintenance build-deploy-part-2-db-dependent'
+      - equal:
+          path: spec.activeDeadlineSeconds
+          value: 1800
 
   - it: switches the command on first deploy
     template: templates/hooks/job-migrate-application.yaml
@@ -35,7 +38,7 @@ tests:
     asserts:
       - matchRegex:
           path: spec.template.spec.containers[0].command[2]
-          pattern: 'sleep 30 && \./phing cluster-first-deploy$'
+          pattern: 'cd /var/www/html && \./phing cluster-first-deploy$'
 
   - it: loads demo data on first deploy when requested
     template: templates/hooks/job-migrate-application.yaml
@@ -77,6 +80,9 @@ tests:
       - matchRegex:
           path: spec.template.spec.containers[0].command[2]
           pattern: 'set -e'
+      - equal:
+          path: spec.activeDeadlineSeconds
+          value: 600
       - matchRegex:
           path: spec.template.spec.containers[0].command[2]
           pattern: './phing maintenance-off\n'
diff --git a/charts/shopsys-app/values.schema.json b/charts/shopsys-app/values.schema.json
index 51d2a9c..66e0d78 100644
--- a/charts/shopsys-app/values.schema.json
+++ b/charts/shopsys-app/values.schema.json
@@ -158,6 +158,20 @@
             "enabled": { "type": "boolean" },
             "loadDemoData": { "type": "boolean" }
           }
+        },
+        "migration": {
+          "type": "object",
+          "properties": {
+            "enabled": { "type": "boolean" },
+            "activeDeadlineSeconds": { "type": "integer", "minimum": 1 }
+          }
+        },
+        "postDeploy": {
+          "type": "object",
+          "properties": {
+            "enabled": { "type": "boolean" },
+            "activeDeadlineSeconds": { "type": "integer", "minimum": 1 }
+          }
         }
       }
     },
diff --git a/charts/shopsys-app/values.yaml b/charts/shopsys-app/values.yaml
index 583efa2..f446bac 100644
--- a/charts/shopsys-app/values.yaml
+++ b/charts/shopsys-app/values.yaml
@@ -353,9 +353,12 @@ deploy:
       continuous: "-verbose db-migrations-count-with-maintenance build-deploy-part-2-db-dependent"
       firstDeploy: "cluster-first-deploy"
       firstDeployWithDemoData: "cluster-first-deploy db-fixtures-demo plugin-demo-data-load friendly-urls-generate domains-urls-replace elasticsearch-export"
+    # Hard bound for the migration Job (seconds); must exceed the longest migration
+    activeDeadlineSeconds: 1800
     resources: {}
   postDeploy:
     enabled: true
+    activeDeadlineSeconds: 600
     resources: {}
   hooks:
     # kubectlImage is the image used by the cron-suspend hook Job; pinned tag on purpose
diff --git a/docs/migrating-from-shopsys-deployment.md b/docs/migrating-from-shopsys-deployment.md
index c2e2d1f..083d7a1 100644
--- a/docs/migrating-from-shopsys-deployment.md
+++ b/docs/migrating-from-shopsys-deployment.md
@@ -84,8 +84,11 @@ Intentional differences of the phase-1 rewrite; everything else is a 1:1 port.
    `domains-urls-hook` copy because pre-install hooks run before regular resources exist.
 8. **Namespace** is created by the wrapper/helmfile, not applied as a manifest. It is named
    `<project.name>-<environment>` (replaces the "PROJECT_NAME must contain a dash" rule).
-9. **`sleep 30`** in the first-deploy migration command is kept verbatim for parity even
-   though the infra release already guarantees readiness.
+9. **The legacy `sleep 30`** in the first-deploy migration command is removed — the
+   shopsys-infra release is installed with `wait: true` before the migration hook runs, so
+   readiness is guaranteed without a delay. The migration and post-deploy Jobs are
+   additionally bounded by `deploy.migration.activeDeadlineSeconds` (default 1800 s) and
+   `deploy.postDeploy.activeDeadlineSeconds` (default 600 s).
 10. **DISPLAY_FINAL_CONFIGURATION** prints one `helmfile template` output instead of two
     kustomize sections.
 11. **`orchestration/kubernetes/` file overrides and the composer `merge` step are gone** —
diff --git a/tests/golden/scenarios/basic-production/expected/continuous.yaml b/tests/golden/scenarios/basic-production/expected/continuous.yaml
index fda309d..62b6bf4 100644
--- a/tests/golden/scenarios/basic-production/expected/continuous.yaml
+++ b/tests/golden/scenarios/basic-production/expected/continuous.yaml
@@ -1742,6 +1742,9 @@ metadata:
     helm.sh/hook-delete-policy: before-hook-creation
 spec:
   backoffLimit: 0
+  # Bound the migration - a stuck Job otherwise blocks the deploy until the global
+  # helmfile timeout and stays running in the cluster
+  activeDeadlineSeconds: 1800
   template:
     spec:
       volumes:
@@ -1810,6 +1813,7 @@ metadata:
     helm.sh/hook-delete-policy: before-hook-creation
 spec:
   backoffLimit: 0
+  activeDeadlineSeconds: 600
   template:
     spec:
       volumes:
diff --git a/tests/golden/scenarios/basic-production/expected/first-deploy-with-demo-data.yaml b/tests/golden/scenarios/basic-production/expected/first-deploy-with-demo-data.yaml
index 4fa723f..565cef4 100644
--- a/tests/golden/scenarios/basic-production/expected/first-deploy-with-demo-data.yaml
+++ b/tests/golden/scenarios/basic-production/expected/first-deploy-with-demo-data.yaml
@@ -1742,6 +1742,9 @@ metadata:
     helm.sh/hook-delete-policy: before-hook-creation
 spec:
   backoffLimit: 0
+  # Bound the migration - a stuck Job otherwise blocks the deploy until the global
+  # helmfile timeout and stays running in the cluster
+  activeDeadlineSeconds: 1800
   template:
     spec:
       volumes:
@@ -1751,7 +1754,7 @@ spec:
       containers:
         - name: migrate-application
           image: "v1.0.0"
-          command: ["sh", "-c", "cd /var/www/html && sleep 30 && ./phing cluster-first-deploy db-fixtures-demo plugin-demo-data-load friendly-urls-generate domains-urls-replace elasticsearch-export"]
+          command: ["sh", "-c", "cd /var/www/html && ./phing cluster-first-deploy db-fixtures-demo plugin-demo-data-load friendly-urls-generate domains-urls-replace elasticsearch-export"]
           envFrom:
             - secretRef:
                 name: app-secret-env-hook
@@ -1810,6 +1813,7 @@ metadata:
     helm.sh/hook-delete-policy: before-hook-creation
 spec:
   backoffLimit: 0
+  activeDeadlineSeconds: 600
   template:
     spec:
       volumes:
diff --git a/tests/golden/scenarios/basic-production/expected/first-deploy.yaml b/tests/golden/scenarios/basic-production/expected/first-deploy.yaml
index 6e59144..49cfd05 100644
--- a/tests/golden/scenarios/basic-production/expected/first-deploy.yaml
+++ b/tests/golden/scenarios/basic-production/expected/first-deploy.yaml
@@ -1742,6 +1742,9 @@ metadata:
     helm.sh/hook-delete-policy: before-hook-creation
 spec:
   backoffLimit: 0
+  # Bound the migration - a stuck Job otherwise blocks the deploy until the global
+  # helmfile timeout and stays running in the cluster
+  activeDeadlineSeconds: 1800
   template:
     spec:
       volumes:
@@ -1751,7 +1754,7 @@ spec:
       containers:
         - name: migrate-application
           image: "v1.0.0"
-          command: ["sh", "-c", "cd /var/www/html && sleep 30 && ./phing cluster-first-deploy"]
+          command: ["sh", "-c", "cd /var/www/html && ./phing cluster-first-deploy"]
           envFrom:
             - secretRef:
                 name: app-secret-env-hook
@@ -1810,6 +1813,7 @@ metadata:
     helm.sh/hook-delete-policy: before-hook-creation
 spec:
   backoffLimit: 0
+  activeDeadlineSeconds: 600
   template:
     spec:
       volumes:
diff --git a/tests/golden/scenarios/development-single-domain/expected/continuous.yaml b/tests/golden/scenarios/development-single-domain/expected/continuous.yaml
index d7a06fd..88393de 100644
--- a/tests/golden/scenarios/development-single-domain/expected/continuous.yaml
+++ b/tests/golden/scenarios/development-single-domain/expected/continuous.yaml
@@ -1519,6 +1519,9 @@ metadata:
     helm.sh/hook-delete-policy: before-hook-creation
 spec:
   backoffLimit: 0
+  # Bound the migration - a stuck Job otherwise blocks the deploy until the global
+  # helmfile timeout and stays running in the cluster
+  activeDeadlineSeconds: 1800
   template:
     spec:
       volumes:
@@ -1587,6 +1590,7 @@ metadata:
     helm.sh/hook-delete-policy: before-hook-creation
 spec:
   backoffLimit: 0
+  activeDeadlineSeconds: 600
   template:
     spec:
       volumes:
diff --git a/tests/golden/scenarios/development-single-domain/expected/first-deploy-with-demo-data.yaml b/tests/golden/scenarios/development-single-domain/expected/first-deploy-with-demo-data.yaml
index f24e535..aaf370d 100644
--- a/tests/golden/scenarios/development-single-domain/expected/first-deploy-with-demo-data.yaml
+++ b/tests/golden/scenarios/development-single-domain/expected/first-deploy-with-demo-data.yaml
@@ -1519,6 +1519,9 @@ metadata:
     helm.sh/hook-delete-policy: before-hook-creation
 spec:
   backoffLimit: 0
+  # Bound the migration - a stuck Job otherwise blocks the deploy until the global
+  # helmfile timeout and stays running in the cluster
+  activeDeadlineSeconds: 1800
   template:
     spec:
       volumes:
@@ -1528,7 +1531,7 @@ spec:
       containers:
         - name: migrate-application
           image: "v1.0.0"
-          command: ["sh", "-c", "cd /var/www/html && sleep 30 && ./phing cluster-first-deploy db-fixtures-demo plugin-demo-data-load friendly-urls-generate domains-urls-replace elasticsearch-export"]
+          command: ["sh", "-c", "cd /var/www/html && ./phing cluster-first-deploy db-fixtures-demo plugin-demo-data-load friendly-urls-generate domains-urls-replace elasticsearch-export"]
           envFrom:
             - secretRef:
                 name: app-secret-env-hook
@@ -1587,6 +1590,7 @@ metadata:
     helm.sh/hook-delete-policy: before-hook-creation
 spec:
   backoffLimit: 0
+  activeDeadlineSeconds: 600
   template:
     spec:
       volumes:
diff --git a/tests/golden/scenarios/development-single-domain/expected/first-deploy.yaml b/tests/golden/scenarios/development-single-domain/expected/first-deploy.yaml
index c742794..dd4ad0b 100644
--- a/tests/golden/scenarios/development-single-domain/expected/first-deploy.yaml
+++ b/tests/golden/scenarios/development-single-domain/expected/first-deploy.yaml
@@ -1519,6 +1519,9 @@ metadata:
     helm.sh/hook-delete-policy: before-hook-creation
 spec:
   backoffLimit: 0
+  # Bound the migration - a stuck Job otherwise blocks the deploy until the global
+  # helmfile timeout and stays running in the cluster
+  activeDeadlineSeconds: 1800
   template:
     spec:
       volumes:
@@ -1528,7 +1531,7 @@ spec:
       containers:
         - name: migrate-application
           image: "v1.0.0"
-          command: ["sh", "-c", "cd /var/www/html && sleep 30 && ./phing cluster-first-deploy"]
+          command: ["sh", "-c", "cd /var/www/html && ./phing cluster-first-deploy"]
           envFrom:
             - secretRef:
                 name: app-secret-env-hook
@@ -1587,6 +1590,7 @@ metadata:
     helm.sh/hook-delete-policy: before-hook-creation
 spec:
   backoffLimit: 0
+  activeDeadlineSeconds: 600
   template:
     spec:
       volumes:
diff --git a/tests/golden/scenarios/development-with-cloudflare/expected/continuous.yaml b/tests/golden/scenarios/development-with-cloudflare/expected/continuous.yaml
index 49a4eb4..88ddb2c 100644
--- a/tests/golden/scenarios/development-with-cloudflare/expected/continuous.yaml
+++ b/tests/golden/scenarios/development-with-cloudflare/expected/continuous.yaml
@@ -1573,6 +1573,9 @@ metadata:
     helm.sh/hook-delete-policy: before-hook-creation
 spec:
   backoffLimit: 0
+  # Bound the migration - a stuck Job otherwise blocks the deploy until the global
+  # helmfile timeout and stays running in the cluster
+  activeDeadlineSeconds: 1800
   template:
     spec:
       volumes:
@@ -1641,6 +1644,7 @@ metadata:
     helm.sh/hook-delete-policy: before-hook-creation
 spec:
   backoffLimit: 0
+  activeDeadlineSeconds: 600
   template:
     spec:
       volumes:
diff --git a/tests/golden/scenarios/development-with-cloudflare/expected/first-deploy-with-demo-data.yaml b/tests/golden/scenarios/development-with-cloudflare/expected/first-deploy-with-demo-data.yaml
index 2fe4f09..f471859 100644
--- a/tests/golden/scenarios/development-with-cloudflare/expected/first-deploy-with-demo-data.yaml
+++ b/tests/golden/scenarios/development-with-cloudflare/expected/first-deploy-with-demo-data.yaml
@@ -1573,6 +1573,9 @@ metadata:
     helm.sh/hook-delete-policy: before-hook-creation
 spec:
   backoffLimit: 0
+  # Bound the migration - a stuck Job otherwise blocks the deploy until the global
+  # helmfile timeout and stays running in the cluster
+  activeDeadlineSeconds: 1800
   template:
     spec:
       volumes:
@@ -1582,7 +1585,7 @@ spec:
       containers:
         - name: migrate-application
           image: "v1.0.0"
-          command: ["sh", "-c", "cd /var/www/html && sleep 30 && ./phing cluster-first-deploy db-fixtures-demo plugin-demo-data-load friendly-urls-generate domains-urls-replace elasticsearch-export"]
+          command: ["sh", "-c", "cd /var/www/html && ./phing cluster-first-deploy db-fixtures-demo plugin-demo-data-load friendly-urls-generate domains-urls-replace elasticsearch-export"]
           envFrom:
             - secretRef:
                 name: app-secret-env-hook
@@ -1641,6 +1644,7 @@ metadata:
     helm.sh/hook-delete-policy: before-hook-creation
 spec:
   backoffLimit: 0
+  activeDeadlineSeconds: 600
   template:
     spec:
       volumes:
diff --git a/tests/golden/scenarios/development-with-cloudflare/expected/first-deploy.yaml b/tests/golden/scenarios/development-with-cloudflare/expected/first-deploy.yaml
index ebcc1ea..18b061c 100644
--- a/tests/golden/scenarios/development-with-cloudflare/expected/first-deploy.yaml
+++ b/tests/golden/scenarios/development-with-cloudflare/expected/first-deploy.yaml
@@ -1573,6 +1573,9 @@ metadata:
     helm.sh/hook-delete-policy: before-hook-creation
 spec:
   backoffLimit: 0
+  # Bound the migration - a stuck Job otherwise blocks the deploy until the global
+  # helmfile timeout and stays running in the cluster
+  activeDeadlineSeconds: 1800
   template:
     spec:
       volumes:
@@ -1582,7 +1585,7 @@ spec:
       containers:
         - name: migrate-application
           image: "v1.0.0"
-          command: ["sh", "-c", "cd /var/www/html && sleep 30 && ./phing cluster-first-deploy"]
+          command: ["sh", "-c", "cd /var/www/html && ./phing cluster-first-deploy"]
           envFrom:
             - secretRef:
                 name: app-secret-env-hook
@@ -1641,6 +1644,7 @@ metadata:
     helm.sh/hook-delete-policy: before-hook-creation
 spec:
   backoffLimit: 0
+  activeDeadlineSeconds: 600
   template:
     spec:
       volumes:
diff --git a/tests/golden/scenarios/escaping-env/expected/continuous.yaml b/tests/golden/scenarios/escaping-env/expected/continuous.yaml
index f7d5168..e6099fe 100644
--- a/tests/golden/scenarios/escaping-env/expected/continuous.yaml
+++ b/tests/golden/scenarios/escaping-env/expected/continuous.yaml
@@ -1769,6 +1769,9 @@ metadata:
     helm.sh/hook-delete-policy: before-hook-creation
 spec:
   backoffLimit: 0
+  # Bound the migration - a stuck Job otherwise blocks the deploy until the global
+  # helmfile timeout and stays running in the cluster
+  activeDeadlineSeconds: 1800
   template:
     spec:
       volumes:
@@ -1839,6 +1842,7 @@ metadata:
     helm.sh/hook-delete-policy: before-hook-creation
 spec:
   backoffLimit: 0
+  activeDeadlineSeconds: 600
   template:
     spec:
       volumes:
diff --git a/tests/golden/scenarios/escaping-env/expected/first-deploy-with-demo-data.yaml b/tests/golden/scenarios/escaping-env/expected/first-deploy-with-demo-data.yaml
index 10aa7d7..62dc8db 100644
--- a/tests/golden/scenarios/escaping-env/expected/first-deploy-with-demo-data.yaml
+++ b/tests/golden/scenarios/escaping-env/expected/first-deploy-with-demo-data.yaml
@@ -1769,6 +1769,9 @@ metadata:
     helm.sh/hook-delete-policy: before-hook-creation
 spec:
   backoffLimit: 0
+  # Bound the migration - a stuck Job otherwise blocks the deploy until the global
+  # helmfile timeout and stays running in the cluster
+  activeDeadlineSeconds: 1800
   template:
     spec:
       volumes:
@@ -1778,7 +1781,7 @@ spec:
       containers:
         - name: migrate-application
           image: "v1.0.0"
-          command: ["sh", "-c", "cd /var/www/html && sleep 30 && ./phing cluster-first-deploy db-fixtures-demo plugin-demo-data-load friendly-urls-generate domains-urls-replace elasticsearch-export"]
+          command: ["sh", "-c", "cd /var/www/html && ./phing cluster-first-deploy db-fixtures-demo plugin-demo-data-load friendly-urls-generate domains-urls-replace elasticsearch-export"]
           envFrom:
             - secretRef:
                 name: app-secret-env-hook
@@ -1839,6 +1842,7 @@ metadata:
     helm.sh/hook-delete-policy: before-hook-creation
 spec:
   backoffLimit: 0
+  activeDeadlineSeconds: 600
   template:
     spec:
       volumes:
diff --git a/tests/golden/scenarios/escaping-env/expected/first-deploy.yaml b/tests/golden/scenarios/escaping-env/expected/first-deploy.yaml
index 050be36..765822f 100644
--- a/tests/golden/scenarios/escaping-env/expected/first-deploy.yaml
+++ b/tests/golden/scenarios/escaping-env/expected/first-deploy.yaml
@@ -1769,6 +1769,9 @@ metadata:
     helm.sh/hook-delete-policy: before-hook-creation
 spec:
   backoffLimit: 0
+  # Bound the migration - a stuck Job otherwise blocks the deploy until the global
+  # helmfile timeout and stays running in the cluster
+  activeDeadlineSeconds: 1800
   template:
     spec:
       volumes:
@@ -1778,7 +1781,7 @@ spec:
       containers:
         - name: migrate-application
           image: "v1.0.0"
-          command: ["sh", "-c", "cd /var/www/html && sleep 30 && ./phing cluster-first-deploy"]
+          command: ["sh", "-c", "cd /var/www/html && ./phing cluster-first-deploy"]
           envFrom:
             - secretRef:
                 name: app-secret-env-hook
@@ -1839,6 +1842,7 @@ metadata:
     helm.sh/hook-delete-policy: before-hook-creation
 spec:
   backoffLimit: 0
+  activeDeadlineSeconds: 600
   template:
     spec:
       volumes:
diff --git a/tests/golden/scenarios/production-with-cloudflare/expected/continuous.yaml b/tests/golden/scenarios/production-with-cloudflare/expected/continuous.yaml
index e15485e..00a720c 100644
--- a/tests/golden/scenarios/production-with-cloudflare/expected/continuous.yaml
+++ b/tests/golden/scenarios/production-with-cloudflare/expected/continuous.yaml
@@ -1874,6 +1874,9 @@ metadata:
     helm.sh/hook-delete-policy: before-hook-creation
 spec:
   backoffLimit: 0
+  # Bound the migration - a stuck Job otherwise blocks the deploy until the global
+  # helmfile timeout and stays running in the cluster
+  activeDeadlineSeconds: 1800
   template:
     spec:
       volumes:
@@ -1942,6 +1945,7 @@ metadata:
     helm.sh/hook-delete-policy: before-hook-creation
 spec:
   backoffLimit: 0
+  activeDeadlineSeconds: 600
   template:
     spec:
       volumes:
diff --git a/tests/golden/scenarios/production-with-cloudflare/expected/first-deploy-with-demo-data.yaml b/tests/golden/scenarios/production-with-cloudflare/expected/first-deploy-with-demo-data.yaml
index ba649cd..55fc7f9 100644
--- a/tests/golden/scenarios/production-with-cloudflare/expected/first-deploy-with-demo-data.yaml
+++ b/tests/golden/scenarios/production-with-cloudflare/expected/first-deploy-with-demo-data.yaml
@@ -1874,6 +1874,9 @@ metadata:
     helm.sh/hook-delete-policy: before-hook-creation
 spec:
   backoffLimit: 0
+  # Bound the migration - a stuck Job otherwise blocks the deploy until the global
+  # helmfile timeout and stays running in the cluster
+  activeDeadlineSeconds: 1800
   template:
     spec:
       volumes:
@@ -1883,7 +1886,7 @@ spec:
       containers:
         - name: migrate-application
           image: "v1.0.0"
-          command: ["sh", "-c", "cd /var/www/html && sleep 30 && ./phing cluster-first-deploy db-fixtures-demo plugin-demo-data-load friendly-urls-generate domains-urls-replace elasticsearch-export"]
+          command: ["sh", "-c", "cd /var/www/html && ./phing cluster-first-deploy db-fixtures-demo plugin-demo-data-load friendly-urls-generate domains-urls-replace elasticsearch-export"]
           envFrom:
             - secretRef:
                 name: app-secret-env-hook
@@ -1942,6 +1945,7 @@ metadata:
     helm.sh/hook-delete-policy: before-hook-creation
 spec:
   backoffLimit: 0
+  activeDeadlineSeconds: 600
   template:
     spec:
       volumes:
diff --git a/tests/golden/scenarios/production-with-cloudflare/expected/first-deploy.yaml b/tests/golden/scenarios/production-with-cloudflare/expected/first-deploy.yaml
index 370b55e..94be774 100644
--- a/tests/golden/scenarios/production-with-cloudflare/expected/first-deploy.yaml
+++ b/tests/golden/scenarios/production-with-cloudflare/expected/first-deploy.yaml
@@ -1874,6 +1874,9 @@ metadata:
     helm.sh/hook-delete-policy: before-hook-creation
 spec:
   backoffLimit: 0
+  # Bound the migration - a stuck Job otherwise blocks the deploy until the global
+  # helmfile timeout and stays running in the cluster
+  activeDeadlineSeconds: 1800
   template:
     spec:
       volumes:
@@ -1883,7 +1886,7 @@ spec:
       containers:
         - name: migrate-application
           image: "v1.0.0"
-          command: ["sh", "-c", "cd /var/www/html && sleep 30 && ./phing cluster-first-deploy"]
+          command: ["sh", "-c", "cd /var/www/html && ./phing cluster-first-deploy"]
           envFrom:
             - secretRef:
                 name: app-secret-env-hook
@@ -1942,6 +1945,7 @@ metadata:
     helm.sh/hook-delete-policy: before-hook-creation
 spec:
   backoffLimit: 0
+  activeDeadlineSeconds: 600
   template:
     spec:
       volumes:

From 44e4ee794e4b953cbe00c659d174e4416ac4770b Mon Sep 17 00:00:00 2001
From: =?UTF-8?q?Jan=20G=C3=B3recki?= <jan.gorecki@shopsys.com>
Date: Thu, 20 Aug 2026 12:49:53 +0200
Subject: [PATCH 2/5] Address review: stale sleep-30 docs, deadline caveats,
 log fallback

Review follow-ups on the sleep-30 removal / hook-bounding PR:

- Purge the remaining stale sleep-30 references the diff left behind:
  the migrate-application template header comment, the deploy-flow.md
  flow diagram and the CLAUDE.md "preserved oddity" example (now points
  at the deviations register instead of a removed example).
- State the readiness guarantee precisely in deviation entry 9: Redis
  readiness is probe-backed, RabbitMQ has no default readinessProbe so
  the infra wait only guarantees a Running broker container.
- Document the activeDeadlineSeconds keys in docs/values.md and expand
  the values.yaml comments: first deploys with demo data may need much
  more than 1800 s, and the value must stay below DEPLOY_TIMEOUT so a
  stuck migration fails as a DeadlineExceeded Job that the wrapper's
  recovery path handles.
- A deadline-killed Job has its pod deleted, so `kubectl logs job/...`
  finds nothing exactly when logs matter most - print_job_logs now
  falls back to `kubectl describe job` (conditions + events).

No rendered-manifest changes (comment-only template edit), golden
snapshots unchanged.

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
---
 CLAUDE.md                                           |  5 +++--
 .../templates/hooks/job-migrate-application.yaml    |  4 ++--
 charts/shopsys-app/values.yaml                      |  6 +++++-
 deploy/deploy.sh                                    |  6 +++++-
 docs/deploy-flow.md                                 |  2 +-
 docs/migrating-from-shopsys-deployment.md           | 13 +++++++++----
 docs/values.md                                      | 12 ++++++++++--
 7 files changed, 35 insertions(+), 13 deletions(-)

diff --git a/CLAUDE.md b/CLAUDE.md
index 03308a4..8069a07 100644
--- a/CLAUDE.md
+++ b/CLAUDE.md
@@ -10,8 +10,9 @@ Kubernetes. It is a **generic, reusable package** consumed by multiple projects.
 
 Phase 1 (done) is a faithful 1:1 port: the deploy **order** and all **end states** of the
 legacy pipeline are preserved. Phase 2 (planned) is manifest modernization — until then,
-manifests intentionally keep their legacy shape (including oddities like the `sleep 30` in
-the first-deploy migration command). Do not "improve" manifest content without being asked.
+manifests intentionally keep their legacy shape; the few intentional exceptions are
+registered in `docs/migrating-from-shopsys-deployment.md`. Do not "improve" manifest
+content without being asked.
 
 **Helm 4 only** — no Helm 3 backward compatibility (e.g. `helm plugin install` uses
 `--verify=false`, which Helm 3 does not know). Keep CI's `HELM_VERSION` in sync with the
diff --git a/charts/shopsys-app/templates/hooks/job-migrate-application.yaml b/charts/shopsys-app/templates/hooks/job-migrate-application.yaml
index 03c2cf5..ae88710 100644
--- a/charts/shopsys-app/templates/hooks/job-migrate-application.yaml
+++ b/charts/shopsys-app/templates/hooks/job-migrate-application.yaml
@@ -9,8 +9,8 @@ it at the start of the next deploy (legacy "Delete previous migration pod" step)
 
 The command variant is selected by deploy.firstDeploy.*:
 - continuous deploy: check migrations + enable maintenance + run db-dependent build phase
-- first deploy: full first deploy of the cluster (the sleep 30 is kept from the legacy
- package to let infrastructure services settle)
+- first deploy: full first deploy of the cluster (the legacy sleep 30 was dropped - the
+ shopsys-infra release is waited on before this hook runs, see the deviations doc)
 - first deploy with demo data: first deploy + demo data import
 */}}
 {{- if .Values.deploy.migration.enabled }}
diff --git a/charts/shopsys-app/values.yaml b/charts/shopsys-app/values.yaml
index f446bac..aa1da2f 100644
--- a/charts/shopsys-app/values.yaml
+++ b/charts/shopsys-app/values.yaml
@@ -353,11 +353,15 @@ deploy:
       continuous: "-verbose db-migrations-count-with-maintenance build-deploy-part-2-db-dependent"
       firstDeploy: "cluster-first-deploy"
       firstDeployWithDemoData: "cluster-first-deploy db-fixtures-demo plugin-demo-data-load friendly-urls-generate domains-urls-replace elasticsearch-export"
-    # Hard bound for the migration Job (seconds); must exceed the longest migration
+    # Hard bound for the migration Job (seconds); must exceed the longest migration -
+    # first deploys with demo data (elasticsearch-export) may need a much higher value.
+    # Keep it below the helmfile timeout (DEPLOY_TIMEOUT, default 2700) so a stuck
+    # migration fails as a DeadlineExceeded Job instead of a Helm client timeout.
     activeDeadlineSeconds: 1800
     resources: {}
   postDeploy:
     enabled: true
+    # Hard bound for the post-deploy Job (seconds)
     activeDeadlineSeconds: 600
     resources: {}
   hooks:
diff --git a/deploy/deploy.sh b/deploy/deploy.sh
index 9b94c07..e216fe7 100755
--- a/deploy/deploy.sh
+++ b/deploy/deploy.sh
@@ -87,7 +87,11 @@ print_job_logs() {
     local job="$1" title="$2"
     if kubectl -n "${NAMESPACE}" get "job/${job}" > /dev/null 2>&1; then
         section_start "${job//-/_}_logs" "${title}"
-        kubectl logs "job/${job}" --namespace="${NAMESPACE}" || true
+        if ! kubectl logs "job/${job}" --namespace="${NAMESPACE}" 2> /dev/null; then
+            # A deadline-killed Job (activeDeadlineSeconds) has its pod deleted, so no
+            # logs exist - fall back to the Job description (conditions + events)
+            kubectl describe "job/${job}" --namespace="${NAMESPACE}" || true
+        fi
         section_end "${job//-/_}_logs"
     fi
 }
diff --git a/docs/deploy-flow.md b/docs/deploy-flow.md
index 7df31f5..3156cc6 100644
--- a/docs/deploy-flow.md
+++ b/docs/deploy-flow.md
@@ -27,7 +27,7 @@ For the legacy behavior itself see [original-deployment.md](original-deployment.
 │           ├─ hook pre-install,pre-upgrade (w10): migrate-application Job
 │           │     continuous:  phing db-migrations-count-with-maintenance
 │           │                  build-deploy-part-2-db-dependent
-│           │     first:       sleep 30 && phing cluster-first-deploy [+ demo targets]
+│           │     first:       phing cluster-first-deploy [+ demo targets]
 │           │     └─ FAILS → whole apply aborts, old release stays
 │           ├─ manifests apply + rollout wait (webserver, storefront, cron on the
 │           │     new image – the `date` label forces a new pod, consumers, ingresses,
diff --git a/docs/migrating-from-shopsys-deployment.md b/docs/migrating-from-shopsys-deployment.md
index 083d7a1..9e89c1a 100644
--- a/docs/migrating-from-shopsys-deployment.md
+++ b/docs/migrating-from-shopsys-deployment.md
@@ -85,10 +85,15 @@ Intentional differences of the phase-1 rewrite; everything else is a 1:1 port.
 8. **Namespace** is created by the wrapper/helmfile, not applied as a manifest. It is named
    `<project.name>-<environment>` (replaces the "PROJECT_NAME must contain a dash" rule).
 9. **The legacy `sleep 30`** in the first-deploy migration command is removed — the
-   shopsys-infra release is installed with `wait: true` before the migration hook runs, so
-   readiness is guaranteed without a delay. The migration and post-deploy Jobs are
-   additionally bounded by `deploy.migration.activeDeadlineSeconds` (default 1800 s) and
-   `deploy.postDeploy.activeDeadlineSeconds` (default 600 s).
+   shopsys-infra release is installed with `wait: true` before the migration hook runs
+   (Redis readiness is probe-backed; RabbitMQ has no default readinessProbe, so the wait
+   guarantees a Running broker container, not yet an accepting broker). The migration and
+   post-deploy Jobs are additionally bounded by `deploy.migration.activeDeadlineSeconds`
+   (default 1800 s — first deploys with demo data may need much more, see
+   [values.md](values.md)) and `deploy.postDeploy.activeDeadlineSeconds` (default 600 s).
+   When a deadline fires, the Job controller deletes the running pod — the Job fails as
+   `DeadlineExceeded` (the wrapper's recovery path still triggers), but the pod logs are
+   gone; the wrapper then falls back to `kubectl describe job` output.
 10. **DISPLAY_FINAL_CONFIGURATION** prints one `helmfile template` output instead of two
     kustomize sections.
 11. **`orchestration/kubernetes/` file overrides and the composer `merge` step are gone** —
diff --git a/docs/values.md b/docs/values.md
index fd69463..d2de7dd 100644
--- a/docs/values.md
+++ b/docs/values.md
@@ -84,8 +84,9 @@ rabbitmq:                    # infra component + auth/persistence/management
 deploy:
   timestamp: ""              # injected by the wrapper (forces a new cron pod)
   firstDeploy: { enabled: false, loadDemoData: false }
-  migration: { enabled, targets: {continuous, firstDeploy, firstDeployWithDemoData}, resources }
-  postDeploy: { enabled, resources }
+  migration: { enabled, targets: {continuous, firstDeploy, firstDeployWithDemoData},
+               resources, activeDeadlineSeconds }   # hard Job bound, default 1800 s
+  postDeploy: { enabled, resources, activeDeadlineSeconds }   # hard Job bound, default 600 s
   hooks: { kubectlImage, serviceAccountName }
 
 extraManifests: []           # raw manifests (rendered through tpl) — escape hatch
@@ -100,6 +101,13 @@ entries take precedence over `envFrom` in Kubernetes — never define the same k
 Lists (e.g. `security.whitelistIps`, `domains`) **replace** the base value when overridden by
 an environment file — they are not merged. Maps merge deeply.
 
+`deploy.migration.activeDeadlineSeconds` must exceed the longest migration variant of the
+project — first deploys with demo data (`elasticsearch-export` on real data volumes) may
+need a much higher value than the 1800 s default. Keep it below the helmfile timeout
+(`DEPLOY_TIMEOUT`, default 2700 s, raise both together) so a stuck migration fails as a
+`DeadlineExceeded` Job — which the wrapper's recovery path handles — instead of the Helm
+client timing out.
+
 ## Legacy env var → values mapping
 
 | Legacy env var | New location |

From b4da8edb507dad464349f4cf55e3e218f4bc6583 Mon Sep 17 00:00:00 2001
From: =?UTF-8?q?Jan=20G=C3=B3recki?= <jan.gorecki@shopsys.com>
Date: Thu, 20 Aug 2026 13:08:10 +0200
Subject: [PATCH 3/5] Raise the migration deadline default to 3600 s

Maintainer decision from the PR review discussion: one hour covers real
first-deploy migrations better than 30 minutes. DEPLOY_TIMEOUT's default
rises to 5400 s in step, preserving the designed failure semantics - the
Job's DeadlineExceeded (handled by the wrapper's recovery path) must fire
before the helmfile client timeout.

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
---
 charts/shopsys-app/tests/hooks_test.yaml                    | 2 +-
 charts/shopsys-app/values.yaml                              | 4 ++--
 docs/deploy-flow.md                                         | 2 +-
 docs/migrating-from-shopsys-deployment.md                   | 4 ++--
 docs/values.md                                              | 6 +++---
 helmfile.yaml.gotmpl                                        | 2 +-
 .../scenarios/basic-production/expected/continuous.yaml     | 2 +-
 .../expected/first-deploy-with-demo-data.yaml               | 2 +-
 .../scenarios/basic-production/expected/first-deploy.yaml   | 2 +-
 .../development-single-domain/expected/continuous.yaml      | 2 +-
 .../expected/first-deploy-with-demo-data.yaml               | 2 +-
 .../development-single-domain/expected/first-deploy.yaml    | 2 +-
 .../development-with-cloudflare/expected/continuous.yaml    | 2 +-
 .../expected/first-deploy-with-demo-data.yaml               | 2 +-
 .../development-with-cloudflare/expected/first-deploy.yaml  | 2 +-
 .../golden/scenarios/escaping-env/expected/continuous.yaml  | 2 +-
 .../escaping-env/expected/first-deploy-with-demo-data.yaml  | 2 +-
 .../scenarios/escaping-env/expected/first-deploy.yaml       | 2 +-
 .../production-with-cloudflare/expected/continuous.yaml     | 2 +-
 .../expected/first-deploy-with-demo-data.yaml               | 2 +-
 .../production-with-cloudflare/expected/first-deploy.yaml   | 2 +-
 21 files changed, 25 insertions(+), 25 deletions(-)

diff --git a/charts/shopsys-app/tests/hooks_test.yaml b/charts/shopsys-app/tests/hooks_test.yaml
index 9867f2c..3457559 100644
--- a/charts/shopsys-app/tests/hooks_test.yaml
+++ b/charts/shopsys-app/tests/hooks_test.yaml
@@ -27,7 +27,7 @@ tests:
           pattern: 'db-migrations-count-with-maintenance build-deploy-part-2-db-dependent'
       - equal:
           path: spec.activeDeadlineSeconds
-          value: 1800
+          value: 3600
 
   - it: switches the command on first deploy
     template: templates/hooks/job-migrate-application.yaml
diff --git a/charts/shopsys-app/values.yaml b/charts/shopsys-app/values.yaml
index aa1da2f..70020fb 100644
--- a/charts/shopsys-app/values.yaml
+++ b/charts/shopsys-app/values.yaml
@@ -355,9 +355,9 @@ deploy:
       firstDeployWithDemoData: "cluster-first-deploy db-fixtures-demo plugin-demo-data-load friendly-urls-generate domains-urls-replace elasticsearch-export"
     # Hard bound for the migration Job (seconds); must exceed the longest migration -
     # first deploys with demo data (elasticsearch-export) may need a much higher value.
-    # Keep it below the helmfile timeout (DEPLOY_TIMEOUT, default 2700) so a stuck
+    # Keep it below the helmfile timeout (DEPLOY_TIMEOUT, default 5400) so a stuck
     # migration fails as a DeadlineExceeded Job instead of a Helm client timeout.
-    activeDeadlineSeconds: 1800
+    activeDeadlineSeconds: 3600
     resources: {}
   postDeploy:
     enabled: true
diff --git a/docs/deploy-flow.md b/docs/deploy-flow.md
index 3156cc6..4646790 100644
--- a/docs/deploy-flow.md
+++ b/docs/deploy-flow.md
@@ -76,7 +76,7 @@ Both are the **same** `helmfile apply`. On the very first deploy:
 | 9 kustomize variant selection | `deploy.firstDeploy.enabled/loadDemoData` values (set from `FIRST_DEPLOY` env by `runtime.yaml.gotmpl`) |
 | 10 DISPLAY_FINAL_CONFIGURATION | wrapper: `helmfile template` in a collapsible section |
 | 11 apply migrate batch | infra release (redis/rabbitmq) + migrate hook Job |
-| 12 wait for the job | `helm --wait-for-jobs` (bounded by `DEPLOY_TIMEOUT`, default 45 min — the legacy loop was unbounded) |
+| 12 wait for the job | `helm --wait-for-jobs` (bounded by `DEPLOY_TIMEOUT`, default 90 min — the legacy loop was unbounded) |
 | 13 migration failure recovery | wrapper failure branch (scale cron back to 1, maintenance-off, logs, slack, exit 1) |
 | 14 apply cron | regular resource in the main apply; the `date` pod label (from `deploy.timestamp`) forces a new pod |
 | 15+18 HPA delete/re-apply dance | HPAs are permanent; the Deployments omit `replicas` when autoscaling is enabled, so there is nothing to fight (and no transient scale-down) |
diff --git a/docs/migrating-from-shopsys-deployment.md b/docs/migrating-from-shopsys-deployment.md
index 9e89c1a..c120079 100644
--- a/docs/migrating-from-shopsys-deployment.md
+++ b/docs/migrating-from-shopsys-deployment.md
@@ -70,7 +70,7 @@ Intentional differences of the phase-1 rewrite; everything else is a 1:1 port.
    `ENABLE_AUTOSCALING` split into independent `webserver.autoscaling.enabled` and
    `storefront.autoscaling.enabled`.
 3. **Bounded waits**: the legacy migration polling loop and cron-watch wait were unbounded;
-   now `DEPLOY_TIMEOUT` (default 45 min) bounds the apply and
+   now `DEPLOY_TIMEOUT` (default 90 min) bounds the apply and
    `cron.terminationGracePeriodSeconds` (default 3600 s) bounds the cron drain.
 4. **Cron locking** happens inside the terminating pod (preStop) instead of `kubectl exec`
    into a pod that stayed alive through the migration. Net effect is identical: no crons run
@@ -89,7 +89,7 @@ Intentional differences of the phase-1 rewrite; everything else is a 1:1 port.
    (Redis readiness is probe-backed; RabbitMQ has no default readinessProbe, so the wait
    guarantees a Running broker container, not yet an accepting broker). The migration and
    post-deploy Jobs are additionally bounded by `deploy.migration.activeDeadlineSeconds`
-   (default 1800 s — first deploys with demo data may need much more, see
+   (default 3600 s — first deploys with demo data may need much more, see
    [values.md](values.md)) and `deploy.postDeploy.activeDeadlineSeconds` (default 600 s).
    When a deadline fires, the Job controller deletes the running pod — the Job fails as
    `DeadlineExceeded` (the wrapper's recovery path still triggers), but the pod logs are
diff --git a/docs/values.md b/docs/values.md
index d2de7dd..b6c04bc 100644
--- a/docs/values.md
+++ b/docs/values.md
@@ -85,7 +85,7 @@ deploy:
   timestamp: ""              # injected by the wrapper (forces a new cron pod)
   firstDeploy: { enabled: false, loadDemoData: false }
   migration: { enabled, targets: {continuous, firstDeploy, firstDeployWithDemoData},
-               resources, activeDeadlineSeconds }   # hard Job bound, default 1800 s
+               resources, activeDeadlineSeconds }   # hard Job bound, default 3600 s
   postDeploy: { enabled, resources, activeDeadlineSeconds }   # hard Job bound, default 600 s
   hooks: { kubectlImage, serviceAccountName }
 
@@ -103,8 +103,8 @@ an environment file — they are not merged. Maps merge deeply.
 
 `deploy.migration.activeDeadlineSeconds` must exceed the longest migration variant of the
 project — first deploys with demo data (`elasticsearch-export` on real data volumes) may
-need a much higher value than the 1800 s default. Keep it below the helmfile timeout
-(`DEPLOY_TIMEOUT`, default 2700 s, raise both together) so a stuck migration fails as a
+need a much higher value than the 3600 s default. Keep it below the helmfile timeout
+(`DEPLOY_TIMEOUT`, default 5400 s, raise both together) so a stuck migration fails as a
 `DeadlineExceeded` Job — which the wrapper's recovery path handles — instead of the Helm
 client timing out.
 
diff --git a/helmfile.yaml.gotmpl b/helmfile.yaml.gotmpl
index 33f19e8..bb48109 100644
--- a/helmfile.yaml.gotmpl
+++ b/helmfile.yaml.gotmpl
@@ -46,7 +46,7 @@ releases:
       - {{ $namespace }}/shopsys-infra
     wait: true
     waitForJobs: true
-    timeout: {{ env "DEPLOY_TIMEOUT" | default 2700 }}
+    timeout: {{ env "DEPLOY_TIMEOUT" | default 5400 }}
     values:
       - {{ toYaml .Values | nindent 8 }}
       - project:
diff --git a/tests/golden/scenarios/basic-production/expected/continuous.yaml b/tests/golden/scenarios/basic-production/expected/continuous.yaml
index 62b6bf4..37456a7 100644
--- a/tests/golden/scenarios/basic-production/expected/continuous.yaml
+++ b/tests/golden/scenarios/basic-production/expected/continuous.yaml
@@ -1744,7 +1744,7 @@ spec:
   backoffLimit: 0
   # Bound the migration - a stuck Job otherwise blocks the deploy until the global
   # helmfile timeout and stays running in the cluster
-  activeDeadlineSeconds: 1800
+  activeDeadlineSeconds: 3600
   template:
     spec:
       volumes:
diff --git a/tests/golden/scenarios/basic-production/expected/first-deploy-with-demo-data.yaml b/tests/golden/scenarios/basic-production/expected/first-deploy-with-demo-data.yaml
index 565cef4..6e78a0e 100644
--- a/tests/golden/scenarios/basic-production/expected/first-deploy-with-demo-data.yaml
+++ b/tests/golden/scenarios/basic-production/expected/first-deploy-with-demo-data.yaml
@@ -1744,7 +1744,7 @@ spec:
   backoffLimit: 0
   # Bound the migration - a stuck Job otherwise blocks the deploy until the global
   # helmfile timeout and stays running in the cluster
-  activeDeadlineSeconds: 1800
+  activeDeadlineSeconds: 3600
   template:
     spec:
       volumes:
diff --git a/tests/golden/scenarios/basic-production/expected/first-deploy.yaml b/tests/golden/scenarios/basic-production/expected/first-deploy.yaml
index 49cfd05..c4f46aa 100644
--- a/tests/golden/scenarios/basic-production/expected/first-deploy.yaml
+++ b/tests/golden/scenarios/basic-production/expected/first-deploy.yaml
@@ -1744,7 +1744,7 @@ spec:
   backoffLimit: 0
   # Bound the migration - a stuck Job otherwise blocks the deploy until the global
   # helmfile timeout and stays running in the cluster
-  activeDeadlineSeconds: 1800
+  activeDeadlineSeconds: 3600
   template:
     spec:
       volumes:
diff --git a/tests/golden/scenarios/development-single-domain/expected/continuous.yaml b/tests/golden/scenarios/development-single-domain/expected/continuous.yaml
index 88393de..0e549ca 100644
--- a/tests/golden/scenarios/development-single-domain/expected/continuous.yaml
+++ b/tests/golden/scenarios/development-single-domain/expected/continuous.yaml
@@ -1521,7 +1521,7 @@ spec:
   backoffLimit: 0
   # Bound the migration - a stuck Job otherwise blocks the deploy until the global
   # helmfile timeout and stays running in the cluster
-  activeDeadlineSeconds: 1800
+  activeDeadlineSeconds: 3600
   template:
     spec:
       volumes:
diff --git a/tests/golden/scenarios/development-single-domain/expected/first-deploy-with-demo-data.yaml b/tests/golden/scenarios/development-single-domain/expected/first-deploy-with-demo-data.yaml
index aaf370d..e998961 100644
--- a/tests/golden/scenarios/development-single-domain/expected/first-deploy-with-demo-data.yaml
+++ b/tests/golden/scenarios/development-single-domain/expected/first-deploy-with-demo-data.yaml
@@ -1521,7 +1521,7 @@ spec:
   backoffLimit: 0
   # Bound the migration - a stuck Job otherwise blocks the deploy until the global
   # helmfile timeout and stays running in the cluster
-  activeDeadlineSeconds: 1800
+  activeDeadlineSeconds: 3600
   template:
     spec:
       volumes:
diff --git a/tests/golden/scenarios/development-single-domain/expected/first-deploy.yaml b/tests/golden/scenarios/development-single-domain/expected/first-deploy.yaml
index dd4ad0b..3d8fc8a 100644
--- a/tests/golden/scenarios/development-single-domain/expected/first-deploy.yaml
+++ b/tests/golden/scenarios/development-single-domain/expected/first-deploy.yaml
@@ -1521,7 +1521,7 @@ spec:
   backoffLimit: 0
   # Bound the migration - a stuck Job otherwise blocks the deploy until the global
   # helmfile timeout and stays running in the cluster
-  activeDeadlineSeconds: 1800
+  activeDeadlineSeconds: 3600
   template:
     spec:
       volumes:
diff --git a/tests/golden/scenarios/development-with-cloudflare/expected/continuous.yaml b/tests/golden/scenarios/development-with-cloudflare/expected/continuous.yaml
index 88ddb2c..b296404 100644
--- a/tests/golden/scenarios/development-with-cloudflare/expected/continuous.yaml
+++ b/tests/golden/scenarios/development-with-cloudflare/expected/continuous.yaml
@@ -1575,7 +1575,7 @@ spec:
   backoffLimit: 0
   # Bound the migration - a stuck Job otherwise blocks the deploy until the global
   # helmfile timeout and stays running in the cluster
-  activeDeadlineSeconds: 1800
+  activeDeadlineSeconds: 3600
   template:
     spec:
       volumes:
diff --git a/tests/golden/scenarios/development-with-cloudflare/expected/first-deploy-with-demo-data.yaml b/tests/golden/scenarios/development-with-cloudflare/expected/first-deploy-with-demo-data.yaml
index f471859..70358cb 100644
--- a/tests/golden/scenarios/development-with-cloudflare/expected/first-deploy-with-demo-data.yaml
+++ b/tests/golden/scenarios/development-with-cloudflare/expected/first-deploy-with-demo-data.yaml
@@ -1575,7 +1575,7 @@ spec:
   backoffLimit: 0
   # Bound the migration - a stuck Job otherwise blocks the deploy until the global
   # helmfile timeout and stays running in the cluster
-  activeDeadlineSeconds: 1800
+  activeDeadlineSeconds: 3600
   template:
     spec:
       volumes:
diff --git a/tests/golden/scenarios/development-with-cloudflare/expected/first-deploy.yaml b/tests/golden/scenarios/development-with-cloudflare/expected/first-deploy.yaml
index 18b061c..b116a73 100644
--- a/tests/golden/scenarios/development-with-cloudflare/expected/first-deploy.yaml
+++ b/tests/golden/scenarios/development-with-cloudflare/expected/first-deploy.yaml
@@ -1575,7 +1575,7 @@ spec:
   backoffLimit: 0
   # Bound the migration - a stuck Job otherwise blocks the deploy until the global
   # helmfile timeout and stays running in the cluster
-  activeDeadlineSeconds: 1800
+  activeDeadlineSeconds: 3600
   template:
     spec:
       volumes:
diff --git a/tests/golden/scenarios/escaping-env/expected/continuous.yaml b/tests/golden/scenarios/escaping-env/expected/continuous.yaml
index e6099fe..2171cea 100644
--- a/tests/golden/scenarios/escaping-env/expected/continuous.yaml
+++ b/tests/golden/scenarios/escaping-env/expected/continuous.yaml
@@ -1771,7 +1771,7 @@ spec:
   backoffLimit: 0
   # Bound the migration - a stuck Job otherwise blocks the deploy until the global
   # helmfile timeout and stays running in the cluster
-  activeDeadlineSeconds: 1800
+  activeDeadlineSeconds: 3600
   template:
     spec:
       volumes:
diff --git a/tests/golden/scenarios/escaping-env/expected/first-deploy-with-demo-data.yaml b/tests/golden/scenarios/escaping-env/expected/first-deploy-with-demo-data.yaml
index 62dc8db..661d58c 100644
--- a/tests/golden/scenarios/escaping-env/expected/first-deploy-with-demo-data.yaml
+++ b/tests/golden/scenarios/escaping-env/expected/first-deploy-with-demo-data.yaml
@@ -1771,7 +1771,7 @@ spec:
   backoffLimit: 0
   # Bound the migration - a stuck Job otherwise blocks the deploy until the global
   # helmfile timeout and stays running in the cluster
-  activeDeadlineSeconds: 1800
+  activeDeadlineSeconds: 3600
   template:
     spec:
       volumes:
diff --git a/tests/golden/scenarios/escaping-env/expected/first-deploy.yaml b/tests/golden/scenarios/escaping-env/expected/first-deploy.yaml
index 765822f..4c92b89 100644
--- a/tests/golden/scenarios/escaping-env/expected/first-deploy.yaml
+++ b/tests/golden/scenarios/escaping-env/expected/first-deploy.yaml
@@ -1771,7 +1771,7 @@ spec:
   backoffLimit: 0
   # Bound the migration - a stuck Job otherwise blocks the deploy until the global
   # helmfile timeout and stays running in the cluster
-  activeDeadlineSeconds: 1800
+  activeDeadlineSeconds: 3600
   template:
     spec:
       volumes:
diff --git a/tests/golden/scenarios/production-with-cloudflare/expected/continuous.yaml b/tests/golden/scenarios/production-with-cloudflare/expected/continuous.yaml
index 00a720c..9447910 100644
--- a/tests/golden/scenarios/production-with-cloudflare/expected/continuous.yaml
+++ b/tests/golden/scenarios/production-with-cloudflare/expected/continuous.yaml
@@ -1876,7 +1876,7 @@ spec:
   backoffLimit: 0
   # Bound the migration - a stuck Job otherwise blocks the deploy until the global
   # helmfile timeout and stays running in the cluster
-  activeDeadlineSeconds: 1800
+  activeDeadlineSeconds: 3600
   template:
     spec:
       volumes:
diff --git a/tests/golden/scenarios/production-with-cloudflare/expected/first-deploy-with-demo-data.yaml b/tests/golden/scenarios/production-with-cloudflare/expected/first-deploy-with-demo-data.yaml
index 55fc7f9..3528ee7 100644
--- a/tests/golden/scenarios/production-with-cloudflare/expected/first-deploy-with-demo-data.yaml
+++ b/tests/golden/scenarios/production-with-cloudflare/expected/first-deploy-with-demo-data.yaml
@@ -1876,7 +1876,7 @@ spec:
   backoffLimit: 0
   # Bound the migration - a stuck Job otherwise blocks the deploy until the global
   # helmfile timeout and stays running in the cluster
-  activeDeadlineSeconds: 1800
+  activeDeadlineSeconds: 3600
   template:
     spec:
       volumes:
diff --git a/tests/golden/scenarios/production-with-cloudflare/expected/first-deploy.yaml b/tests/golden/scenarios/production-with-cloudflare/expected/first-deploy.yaml
index 94be774..6956794 100644
--- a/tests/golden/scenarios/production-with-cloudflare/expected/first-deploy.yaml
+++ b/tests/golden/scenarios/production-with-cloudflare/expected/first-deploy.yaml
@@ -1876,7 +1876,7 @@ spec:
   backoffLimit: 0
   # Bound the migration - a stuck Job otherwise blocks the deploy until the global
   # helmfile timeout and stays running in the cluster
-  activeDeadlineSeconds: 1800
+  activeDeadlineSeconds: 3600
   template:
     spec:
       volumes:

From f802912069246b8dc72f1163ae99fb5a76e6d971 Mon Sep 17 00:00:00 2001
From: =?UTF-8?q?Jan=20G=C3=B3recki?= <jan.gorecki@shopsys.com>
Date: Thu, 20 Aug 2026 13:13:08 +0200
Subject: [PATCH 4/5] Revert "Raise the migration deadline default to 3600 s"

This reverts commit b4da8edb507dad464349f4cf55e3e218f4bc6583.
---
 charts/shopsys-app/tests/hooks_test.yaml                    | 2 +-
 charts/shopsys-app/values.yaml                              | 4 ++--
 docs/deploy-flow.md                                         | 2 +-
 docs/migrating-from-shopsys-deployment.md                   | 4 ++--
 docs/values.md                                              | 6 +++---
 helmfile.yaml.gotmpl                                        | 2 +-
 .../scenarios/basic-production/expected/continuous.yaml     | 2 +-
 .../expected/first-deploy-with-demo-data.yaml               | 2 +-
 .../scenarios/basic-production/expected/first-deploy.yaml   | 2 +-
 .../development-single-domain/expected/continuous.yaml      | 2 +-
 .../expected/first-deploy-with-demo-data.yaml               | 2 +-
 .../development-single-domain/expected/first-deploy.yaml    | 2 +-
 .../development-with-cloudflare/expected/continuous.yaml    | 2 +-
 .../expected/first-deploy-with-demo-data.yaml               | 2 +-
 .../development-with-cloudflare/expected/first-deploy.yaml  | 2 +-
 .../golden/scenarios/escaping-env/expected/continuous.yaml  | 2 +-
 .../escaping-env/expected/first-deploy-with-demo-data.yaml  | 2 +-
 .../scenarios/escaping-env/expected/first-deploy.yaml       | 2 +-
 .../production-with-cloudflare/expected/continuous.yaml     | 2 +-
 .../expected/first-deploy-with-demo-data.yaml               | 2 +-
 .../production-with-cloudflare/expected/first-deploy.yaml   | 2 +-
 21 files changed, 25 insertions(+), 25 deletions(-)

diff --git a/charts/shopsys-app/tests/hooks_test.yaml b/charts/shopsys-app/tests/hooks_test.yaml
index 3457559..9867f2c 100644
--- a/charts/shopsys-app/tests/hooks_test.yaml
+++ b/charts/shopsys-app/tests/hooks_test.yaml
@@ -27,7 +27,7 @@ tests:
           pattern: 'db-migrations-count-with-maintenance build-deploy-part-2-db-dependent'
       - equal:
           path: spec.activeDeadlineSeconds
-          value: 3600
+          value: 1800
 
   - it: switches the command on first deploy
     template: templates/hooks/job-migrate-application.yaml
diff --git a/charts/shopsys-app/values.yaml b/charts/shopsys-app/values.yaml
index 70020fb..aa1da2f 100644
--- a/charts/shopsys-app/values.yaml
+++ b/charts/shopsys-app/values.yaml
@@ -355,9 +355,9 @@ deploy:
       firstDeployWithDemoData: "cluster-first-deploy db-fixtures-demo plugin-demo-data-load friendly-urls-generate domains-urls-replace elasticsearch-export"
     # Hard bound for the migration Job (seconds); must exceed the longest migration -
     # first deploys with demo data (elasticsearch-export) may need a much higher value.
-    # Keep it below the helmfile timeout (DEPLOY_TIMEOUT, default 5400) so a stuck
+    # Keep it below the helmfile timeout (DEPLOY_TIMEOUT, default 2700) so a stuck
     # migration fails as a DeadlineExceeded Job instead of a Helm client timeout.
-    activeDeadlineSeconds: 3600
+    activeDeadlineSeconds: 1800
     resources: {}
   postDeploy:
     enabled: true
diff --git a/docs/deploy-flow.md b/docs/deploy-flow.md
index 4646790..3156cc6 100644
--- a/docs/deploy-flow.md
+++ b/docs/deploy-flow.md
@@ -76,7 +76,7 @@ Both are the **same** `helmfile apply`. On the very first deploy:
 | 9 kustomize variant selection | `deploy.firstDeploy.enabled/loadDemoData` values (set from `FIRST_DEPLOY` env by `runtime.yaml.gotmpl`) |
 | 10 DISPLAY_FINAL_CONFIGURATION | wrapper: `helmfile template` in a collapsible section |
 | 11 apply migrate batch | infra release (redis/rabbitmq) + migrate hook Job |
-| 12 wait for the job | `helm --wait-for-jobs` (bounded by `DEPLOY_TIMEOUT`, default 90 min — the legacy loop was unbounded) |
+| 12 wait for the job | `helm --wait-for-jobs` (bounded by `DEPLOY_TIMEOUT`, default 45 min — the legacy loop was unbounded) |
 | 13 migration failure recovery | wrapper failure branch (scale cron back to 1, maintenance-off, logs, slack, exit 1) |
 | 14 apply cron | regular resource in the main apply; the `date` pod label (from `deploy.timestamp`) forces a new pod |
 | 15+18 HPA delete/re-apply dance | HPAs are permanent; the Deployments omit `replicas` when autoscaling is enabled, so there is nothing to fight (and no transient scale-down) |
diff --git a/docs/migrating-from-shopsys-deployment.md b/docs/migrating-from-shopsys-deployment.md
index c120079..9e89c1a 100644
--- a/docs/migrating-from-shopsys-deployment.md
+++ b/docs/migrating-from-shopsys-deployment.md
@@ -70,7 +70,7 @@ Intentional differences of the phase-1 rewrite; everything else is a 1:1 port.
    `ENABLE_AUTOSCALING` split into independent `webserver.autoscaling.enabled` and
    `storefront.autoscaling.enabled`.
 3. **Bounded waits**: the legacy migration polling loop and cron-watch wait were unbounded;
-   now `DEPLOY_TIMEOUT` (default 90 min) bounds the apply and
+   now `DEPLOY_TIMEOUT` (default 45 min) bounds the apply and
    `cron.terminationGracePeriodSeconds` (default 3600 s) bounds the cron drain.
 4. **Cron locking** happens inside the terminating pod (preStop) instead of `kubectl exec`
    into a pod that stayed alive through the migration. Net effect is identical: no crons run
@@ -89,7 +89,7 @@ Intentional differences of the phase-1 rewrite; everything else is a 1:1 port.
    (Redis readiness is probe-backed; RabbitMQ has no default readinessProbe, so the wait
    guarantees a Running broker container, not yet an accepting broker). The migration and
    post-deploy Jobs are additionally bounded by `deploy.migration.activeDeadlineSeconds`
-   (default 3600 s — first deploys with demo data may need much more, see
+   (default 1800 s — first deploys with demo data may need much more, see
    [values.md](values.md)) and `deploy.postDeploy.activeDeadlineSeconds` (default 600 s).
    When a deadline fires, the Job controller deletes the running pod — the Job fails as
    `DeadlineExceeded` (the wrapper's recovery path still triggers), but the pod logs are
diff --git a/docs/values.md b/docs/values.md
index b6c04bc..d2de7dd 100644
--- a/docs/values.md
+++ b/docs/values.md
@@ -85,7 +85,7 @@ deploy:
   timestamp: ""              # injected by the wrapper (forces a new cron pod)
   firstDeploy: { enabled: false, loadDemoData: false }
   migration: { enabled, targets: {continuous, firstDeploy, firstDeployWithDemoData},
-               resources, activeDeadlineSeconds }   # hard Job bound, default 3600 s
+               resources, activeDeadlineSeconds }   # hard Job bound, default 1800 s
   postDeploy: { enabled, resources, activeDeadlineSeconds }   # hard Job bound, default 600 s
   hooks: { kubectlImage, serviceAccountName }
 
@@ -103,8 +103,8 @@ an environment file — they are not merged. Maps merge deeply.
 
 `deploy.migration.activeDeadlineSeconds` must exceed the longest migration variant of the
 project — first deploys with demo data (`elasticsearch-export` on real data volumes) may
-need a much higher value than the 3600 s default. Keep it below the helmfile timeout
-(`DEPLOY_TIMEOUT`, default 5400 s, raise both together) so a stuck migration fails as a
+need a much higher value than the 1800 s default. Keep it below the helmfile timeout
+(`DEPLOY_TIMEOUT`, default 2700 s, raise both together) so a stuck migration fails as a
 `DeadlineExceeded` Job — which the wrapper's recovery path handles — instead of the Helm
 client timing out.
 
diff --git a/helmfile.yaml.gotmpl b/helmfile.yaml.gotmpl
index bb48109..33f19e8 100644
--- a/helmfile.yaml.gotmpl
+++ b/helmfile.yaml.gotmpl
@@ -46,7 +46,7 @@ releases:
       - {{ $namespace }}/shopsys-infra
     wait: true
     waitForJobs: true
-    timeout: {{ env "DEPLOY_TIMEOUT" | default 5400 }}
+    timeout: {{ env "DEPLOY_TIMEOUT" | default 2700 }}
     values:
       - {{ toYaml .Values | nindent 8 }}
       - project:
diff --git a/tests/golden/scenarios/basic-production/expected/continuous.yaml b/tests/golden/scenarios/basic-production/expected/continuous.yaml
index 37456a7..62b6bf4 100644
--- a/tests/golden/scenarios/basic-production/expected/continuous.yaml
+++ b/tests/golden/scenarios/basic-production/expected/continuous.yaml
@@ -1744,7 +1744,7 @@ spec:
   backoffLimit: 0
   # Bound the migration - a stuck Job otherwise blocks the deploy until the global
   # helmfile timeout and stays running in the cluster
-  activeDeadlineSeconds: 3600
+  activeDeadlineSeconds: 1800
   template:
     spec:
       volumes:
diff --git a/tests/golden/scenarios/basic-production/expected/first-deploy-with-demo-data.yaml b/tests/golden/scenarios/basic-production/expected/first-deploy-with-demo-data.yaml
index 6e78a0e..565cef4 100644
--- a/tests/golden/scenarios/basic-production/expected/first-deploy-with-demo-data.yaml
+++ b/tests/golden/scenarios/basic-production/expected/first-deploy-with-demo-data.yaml
@@ -1744,7 +1744,7 @@ spec:
   backoffLimit: 0
   # Bound the migration - a stuck Job otherwise blocks the deploy until the global
   # helmfile timeout and stays running in the cluster
-  activeDeadlineSeconds: 3600
+  activeDeadlineSeconds: 1800
   template:
     spec:
       volumes:
diff --git a/tests/golden/scenarios/basic-production/expected/first-deploy.yaml b/tests/golden/scenarios/basic-production/expected/first-deploy.yaml
index c4f46aa..49cfd05 100644
--- a/tests/golden/scenarios/basic-production/expected/first-deploy.yaml
+++ b/tests/golden/scenarios/basic-production/expected/first-deploy.yaml
@@ -1744,7 +1744,7 @@ spec:
   backoffLimit: 0
   # Bound the migration - a stuck Job otherwise blocks the deploy until the global
   # helmfile timeout and stays running in the cluster
-  activeDeadlineSeconds: 3600
+  activeDeadlineSeconds: 1800
   template:
     spec:
       volumes:
diff --git a/tests/golden/scenarios/development-single-domain/expected/continuous.yaml b/tests/golden/scenarios/development-single-domain/expected/continuous.yaml
index 0e549ca..88393de 100644
--- a/tests/golden/scenarios/development-single-domain/expected/continuous.yaml
+++ b/tests/golden/scenarios/development-single-domain/expected/continuous.yaml
@@ -1521,7 +1521,7 @@ spec:
   backoffLimit: 0
   # Bound the migration - a stuck Job otherwise blocks the deploy until the global
   # helmfile timeout and stays running in the cluster
-  activeDeadlineSeconds: 3600
+  activeDeadlineSeconds: 1800
   template:
     spec:
       volumes:
diff --git a/tests/golden/scenarios/development-single-domain/expected/first-deploy-with-demo-data.yaml b/tests/golden/scenarios/development-single-domain/expected/first-deploy-with-demo-data.yaml
index e998961..aaf370d 100644
--- a/tests/golden/scenarios/development-single-domain/expected/first-deploy-with-demo-data.yaml
+++ b/tests/golden/scenarios/development-single-domain/expected/first-deploy-with-demo-data.yaml
@@ -1521,7 +1521,7 @@ spec:
   backoffLimit: 0
   # Bound the migration - a stuck Job otherwise blocks the deploy until the global
   # helmfile timeout and stays running in the cluster
-  activeDeadlineSeconds: 3600
+  activeDeadlineSeconds: 1800
   template:
     spec:
       volumes:
diff --git a/tests/golden/scenarios/development-single-domain/expected/first-deploy.yaml b/tests/golden/scenarios/development-single-domain/expected/first-deploy.yaml
index 3d8fc8a..dd4ad0b 100644
--- a/tests/golden/scenarios/development-single-domain/expected/first-deploy.yaml
+++ b/tests/golden/scenarios/development-single-domain/expected/first-deploy.yaml
@@ -1521,7 +1521,7 @@ spec:
   backoffLimit: 0
   # Bound the migration - a stuck Job otherwise blocks the deploy until the global
   # helmfile timeout and stays running in the cluster
-  activeDeadlineSeconds: 3600
+  activeDeadlineSeconds: 1800
   template:
     spec:
       volumes:
diff --git a/tests/golden/scenarios/development-with-cloudflare/expected/continuous.yaml b/tests/golden/scenarios/development-with-cloudflare/expected/continuous.yaml
index b296404..88ddb2c 100644
--- a/tests/golden/scenarios/development-with-cloudflare/expected/continuous.yaml
+++ b/tests/golden/scenarios/development-with-cloudflare/expected/continuous.yaml
@@ -1575,7 +1575,7 @@ spec:
   backoffLimit: 0
   # Bound the migration - a stuck Job otherwise blocks the deploy until the global
   # helmfile timeout and stays running in the cluster
-  activeDeadlineSeconds: 3600
+  activeDeadlineSeconds: 1800
   template:
     spec:
       volumes:
diff --git a/tests/golden/scenarios/development-with-cloudflare/expected/first-deploy-with-demo-data.yaml b/tests/golden/scenarios/development-with-cloudflare/expected/first-deploy-with-demo-data.yaml
index 70358cb..f471859 100644
--- a/tests/golden/scenarios/development-with-cloudflare/expected/first-deploy-with-demo-data.yaml
+++ b/tests/golden/scenarios/development-with-cloudflare/expected/first-deploy-with-demo-data.yaml
@@ -1575,7 +1575,7 @@ spec:
   backoffLimit: 0
   # Bound the migration - a stuck Job otherwise blocks the deploy until the global
   # helmfile timeout and stays running in the cluster
-  activeDeadlineSeconds: 3600
+  activeDeadlineSeconds: 1800
   template:
     spec:
       volumes:
diff --git a/tests/golden/scenarios/development-with-cloudflare/expected/first-deploy.yaml b/tests/golden/scenarios/development-with-cloudflare/expected/first-deploy.yaml
index b116a73..18b061c 100644
--- a/tests/golden/scenarios/development-with-cloudflare/expected/first-deploy.yaml
+++ b/tests/golden/scenarios/development-with-cloudflare/expected/first-deploy.yaml
@@ -1575,7 +1575,7 @@ spec:
   backoffLimit: 0
   # Bound the migration - a stuck Job otherwise blocks the deploy until the global
   # helmfile timeout and stays running in the cluster
-  activeDeadlineSeconds: 3600
+  activeDeadlineSeconds: 1800
   template:
     spec:
       volumes:
diff --git a/tests/golden/scenarios/escaping-env/expected/continuous.yaml b/tests/golden/scenarios/escaping-env/expected/continuous.yaml
index 2171cea..e6099fe 100644
--- a/tests/golden/scenarios/escaping-env/expected/continuous.yaml
+++ b/tests/golden/scenarios/escaping-env/expected/continuous.yaml
@@ -1771,7 +1771,7 @@ spec:
   backoffLimit: 0
   # Bound the migration - a stuck Job otherwise blocks the deploy until the global
   # helmfile timeout and stays running in the cluster
-  activeDeadlineSeconds: 3600
+  activeDeadlineSeconds: 1800
   template:
     spec:
       volumes:
diff --git a/tests/golden/scenarios/escaping-env/expected/first-deploy-with-demo-data.yaml b/tests/golden/scenarios/escaping-env/expected/first-deploy-with-demo-data.yaml
index 661d58c..62dc8db 100644
--- a/tests/golden/scenarios/escaping-env/expected/first-deploy-with-demo-data.yaml
+++ b/tests/golden/scenarios/escaping-env/expected/first-deploy-with-demo-data.yaml
@@ -1771,7 +1771,7 @@ spec:
   backoffLimit: 0
   # Bound the migration - a stuck Job otherwise blocks the deploy until the global
   # helmfile timeout and stays running in the cluster
-  activeDeadlineSeconds: 3600
+  activeDeadlineSeconds: 1800
   template:
     spec:
       volumes:
diff --git a/tests/golden/scenarios/escaping-env/expected/first-deploy.yaml b/tests/golden/scenarios/escaping-env/expected/first-deploy.yaml
index 4c92b89..765822f 100644
--- a/tests/golden/scenarios/escaping-env/expected/first-deploy.yaml
+++ b/tests/golden/scenarios/escaping-env/expected/first-deploy.yaml
@@ -1771,7 +1771,7 @@ spec:
   backoffLimit: 0
   # Bound the migration - a stuck Job otherwise blocks the deploy until the global
   # helmfile timeout and stays running in the cluster
-  activeDeadlineSeconds: 3600
+  activeDeadlineSeconds: 1800
   template:
     spec:
       volumes:
diff --git a/tests/golden/scenarios/production-with-cloudflare/expected/continuous.yaml b/tests/golden/scenarios/production-with-cloudflare/expected/continuous.yaml
index 9447910..00a720c 100644
--- a/tests/golden/scenarios/production-with-cloudflare/expected/continuous.yaml
+++ b/tests/golden/scenarios/production-with-cloudflare/expected/continuous.yaml
@@ -1876,7 +1876,7 @@ spec:
   backoffLimit: 0
   # Bound the migration - a stuck Job otherwise blocks the deploy until the global
   # helmfile timeout and stays running in the cluster
-  activeDeadlineSeconds: 3600
+  activeDeadlineSeconds: 1800
   template:
     spec:
       volumes:
diff --git a/tests/golden/scenarios/production-with-cloudflare/expected/first-deploy-with-demo-data.yaml b/tests/golden/scenarios/production-with-cloudflare/expected/first-deploy-with-demo-data.yaml
index 3528ee7..55fc7f9 100644
--- a/tests/golden/scenarios/production-with-cloudflare/expected/first-deploy-with-demo-data.yaml
+++ b/tests/golden/scenarios/production-with-cloudflare/expected/first-deploy-with-demo-data.yaml
@@ -1876,7 +1876,7 @@ spec:
   backoffLimit: 0
   # Bound the migration - a stuck Job otherwise blocks the deploy until the global
   # helmfile timeout and stays running in the cluster
-  activeDeadlineSeconds: 3600
+  activeDeadlineSeconds: 1800
   template:
     spec:
       volumes:
diff --git a/tests/golden/scenarios/production-with-cloudflare/expected/first-deploy.yaml b/tests/golden/scenarios/production-with-cloudflare/expected/first-deploy.yaml
index 6956794..94be774 100644
--- a/tests/golden/scenarios/production-with-cloudflare/expected/first-deploy.yaml
+++ b/tests/golden/scenarios/production-with-cloudflare/expected/first-deploy.yaml
@@ -1876,7 +1876,7 @@ spec:
   backoffLimit: 0
   # Bound the migration - a stuck Job otherwise blocks the deploy until the global
   # helmfile timeout and stays running in the cluster
-  activeDeadlineSeconds: 3600
+  activeDeadlineSeconds: 1800
   template:
     spec:
       volumes:

From aee673cd45da38ac9808efe51f5af9a81deb8b75 Mon Sep 17 00:00:00 2001
From: =?UTF-8?q?Jan=20G=C3=B3recki?= <jan.gorecki@shopsys.com>
Date: Thu, 20 Aug 2026 13:24:26 +0200
Subject: [PATCH 5/5] Raise the post-deploy Job deadline default to 1800 s

Maintainer decision: post-deploy targets (cache cleanups, part-3) can
legitimately run long. The two hook bounds together can now exceed
DEPLOY_TIMEOUT - acceptable, both hitting their ceiling in one deploy is
not a normal scenario.

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
---
 charts/shopsys-app/tests/hooks_test.yaml                    | 2 +-
 charts/shopsys-app/values.yaml                              | 6 ++++--
 docs/migrating-from-shopsys-deployment.md                   | 4 +++-
 docs/values.md                                              | 2 +-
 .../scenarios/basic-production/expected/continuous.yaml     | 2 +-
 .../expected/first-deploy-with-demo-data.yaml               | 2 +-
 .../scenarios/basic-production/expected/first-deploy.yaml   | 2 +-
 .../development-single-domain/expected/continuous.yaml      | 2 +-
 .../expected/first-deploy-with-demo-data.yaml               | 2 +-
 .../development-single-domain/expected/first-deploy.yaml    | 2 +-
 .../development-with-cloudflare/expected/continuous.yaml    | 2 +-
 .../expected/first-deploy-with-demo-data.yaml               | 2 +-
 .../development-with-cloudflare/expected/first-deploy.yaml  | 2 +-
 .../golden/scenarios/escaping-env/expected/continuous.yaml  | 2 +-
 .../escaping-env/expected/first-deploy-with-demo-data.yaml  | 2 +-
 .../scenarios/escaping-env/expected/first-deploy.yaml       | 2 +-
 .../production-with-cloudflare/expected/continuous.yaml     | 2 +-
 .../expected/first-deploy-with-demo-data.yaml               | 2 +-
 .../production-with-cloudflare/expected/first-deploy.yaml   | 2 +-
 19 files changed, 24 insertions(+), 20 deletions(-)

diff --git a/charts/shopsys-app/tests/hooks_test.yaml b/charts/shopsys-app/tests/hooks_test.yaml
index 9867f2c..3fae5ae 100644
--- a/charts/shopsys-app/tests/hooks_test.yaml
+++ b/charts/shopsys-app/tests/hooks_test.yaml
@@ -82,7 +82,7 @@ tests:
           pattern: 'set -e'
       - equal:
           path: spec.activeDeadlineSeconds
-          value: 600
+          value: 1800
       - matchRegex:
           path: spec.template.spec.containers[0].command[2]
           pattern: './phing maintenance-off\n'
diff --git a/charts/shopsys-app/values.yaml b/charts/shopsys-app/values.yaml
index aa1da2f..71c801d 100644
--- a/charts/shopsys-app/values.yaml
+++ b/charts/shopsys-app/values.yaml
@@ -361,8 +361,10 @@ deploy:
     resources: {}
   postDeploy:
     enabled: true
-    # Hard bound for the post-deploy Job (seconds)
-    activeDeadlineSeconds: 600
+    # Hard bound for the post-deploy Job (seconds). Together with the migration bound this
+    # can exceed DEPLOY_TIMEOUT (2700 s) - acceptable, both hitting their ceiling in one
+    # deploy is not a normal scenario.
+    activeDeadlineSeconds: 1800
     resources: {}
   hooks:
     # kubectlImage is the image used by the cron-suspend hook Job; pinned tag on purpose
diff --git a/docs/migrating-from-shopsys-deployment.md b/docs/migrating-from-shopsys-deployment.md
index 9e89c1a..1a5e6d9 100644
--- a/docs/migrating-from-shopsys-deployment.md
+++ b/docs/migrating-from-shopsys-deployment.md
@@ -90,7 +90,9 @@ Intentional differences of the phase-1 rewrite; everything else is a 1:1 port.
    guarantees a Running broker container, not yet an accepting broker). The migration and
    post-deploy Jobs are additionally bounded by `deploy.migration.activeDeadlineSeconds`
    (default 1800 s — first deploys with demo data may need much more, see
-   [values.md](values.md)) and `deploy.postDeploy.activeDeadlineSeconds` (default 600 s).
+   [values.md](values.md)) and `deploy.postDeploy.activeDeadlineSeconds` (default 1800 s;
+   the two bounds together can exceed `DEPLOY_TIMEOUT` — acceptable, both hitting their
+   ceiling in one deploy is not a normal scenario).
    When a deadline fires, the Job controller deletes the running pod — the Job fails as
    `DeadlineExceeded` (the wrapper's recovery path still triggers), but the pod logs are
    gone; the wrapper then falls back to `kubectl describe job` output.
diff --git a/docs/values.md b/docs/values.md
index d2de7dd..ddd0f12 100644
--- a/docs/values.md
+++ b/docs/values.md
@@ -86,7 +86,7 @@ deploy:
   firstDeploy: { enabled: false, loadDemoData: false }
   migration: { enabled, targets: {continuous, firstDeploy, firstDeployWithDemoData},
                resources, activeDeadlineSeconds }   # hard Job bound, default 1800 s
-  postDeploy: { enabled, resources, activeDeadlineSeconds }   # hard Job bound, default 600 s
+  postDeploy: { enabled, resources, activeDeadlineSeconds }   # hard Job bound, default 1800 s
   hooks: { kubectlImage, serviceAccountName }
 
 extraManifests: []           # raw manifests (rendered through tpl) — escape hatch
diff --git a/tests/golden/scenarios/basic-production/expected/continuous.yaml b/tests/golden/scenarios/basic-production/expected/continuous.yaml
index 62b6bf4..753e900 100644
--- a/tests/golden/scenarios/basic-production/expected/continuous.yaml
+++ b/tests/golden/scenarios/basic-production/expected/continuous.yaml
@@ -1813,7 +1813,7 @@ metadata:
     helm.sh/hook-delete-policy: before-hook-creation
 spec:
   backoffLimit: 0
-  activeDeadlineSeconds: 600
+  activeDeadlineSeconds: 1800
   template:
     spec:
       volumes:
diff --git a/tests/golden/scenarios/basic-production/expected/first-deploy-with-demo-data.yaml b/tests/golden/scenarios/basic-production/expected/first-deploy-with-demo-data.yaml
index 565cef4..5b12854 100644
--- a/tests/golden/scenarios/basic-production/expected/first-deploy-with-demo-data.yaml
+++ b/tests/golden/scenarios/basic-production/expected/first-deploy-with-demo-data.yaml
@@ -1813,7 +1813,7 @@ metadata:
     helm.sh/hook-delete-policy: before-hook-creation
 spec:
   backoffLimit: 0
-  activeDeadlineSeconds: 600
+  activeDeadlineSeconds: 1800
   template:
     spec:
       volumes:
diff --git a/tests/golden/scenarios/basic-production/expected/first-deploy.yaml b/tests/golden/scenarios/basic-production/expected/first-deploy.yaml
index 49cfd05..5db65ff 100644
--- a/tests/golden/scenarios/basic-production/expected/first-deploy.yaml
+++ b/tests/golden/scenarios/basic-production/expected/first-deploy.yaml
@@ -1813,7 +1813,7 @@ metadata:
     helm.sh/hook-delete-policy: before-hook-creation
 spec:
   backoffLimit: 0
-  activeDeadlineSeconds: 600
+  activeDeadlineSeconds: 1800
   template:
     spec:
       volumes:
diff --git a/tests/golden/scenarios/development-single-domain/expected/continuous.yaml b/tests/golden/scenarios/development-single-domain/expected/continuous.yaml
index 88393de..0caab72 100644
--- a/tests/golden/scenarios/development-single-domain/expected/continuous.yaml
+++ b/tests/golden/scenarios/development-single-domain/expected/continuous.yaml
@@ -1590,7 +1590,7 @@ metadata:
     helm.sh/hook-delete-policy: before-hook-creation
 spec:
   backoffLimit: 0
-  activeDeadlineSeconds: 600
+  activeDeadlineSeconds: 1800
   template:
     spec:
       volumes:
diff --git a/tests/golden/scenarios/development-single-domain/expected/first-deploy-with-demo-data.yaml b/tests/golden/scenarios/development-single-domain/expected/first-deploy-with-demo-data.yaml
index aaf370d..8419d4b 100644
--- a/tests/golden/scenarios/development-single-domain/expected/first-deploy-with-demo-data.yaml
+++ b/tests/golden/scenarios/development-single-domain/expected/first-deploy-with-demo-data.yaml
@@ -1590,7 +1590,7 @@ metadata:
     helm.sh/hook-delete-policy: before-hook-creation
 spec:
   backoffLimit: 0
-  activeDeadlineSeconds: 600
+  activeDeadlineSeconds: 1800
   template:
     spec:
       volumes:
diff --git a/tests/golden/scenarios/development-single-domain/expected/first-deploy.yaml b/tests/golden/scenarios/development-single-domain/expected/first-deploy.yaml
index dd4ad0b..7ec9cd1 100644
--- a/tests/golden/scenarios/development-single-domain/expected/first-deploy.yaml
+++ b/tests/golden/scenarios/development-single-domain/expected/first-deploy.yaml
@@ -1590,7 +1590,7 @@ metadata:
     helm.sh/hook-delete-policy: before-hook-creation
 spec:
   backoffLimit: 0
-  activeDeadlineSeconds: 600
+  activeDeadlineSeconds: 1800
   template:
     spec:
       volumes:
diff --git a/tests/golden/scenarios/development-with-cloudflare/expected/continuous.yaml b/tests/golden/scenarios/development-with-cloudflare/expected/continuous.yaml
index 88ddb2c..fa3dc8b 100644
--- a/tests/golden/scenarios/development-with-cloudflare/expected/continuous.yaml
+++ b/tests/golden/scenarios/development-with-cloudflare/expected/continuous.yaml
@@ -1644,7 +1644,7 @@ metadata:
     helm.sh/hook-delete-policy: before-hook-creation
 spec:
   backoffLimit: 0
-  activeDeadlineSeconds: 600
+  activeDeadlineSeconds: 1800
   template:
     spec:
       volumes:
diff --git a/tests/golden/scenarios/development-with-cloudflare/expected/first-deploy-with-demo-data.yaml b/tests/golden/scenarios/development-with-cloudflare/expected/first-deploy-with-demo-data.yaml
index f471859..8aaac29 100644
--- a/tests/golden/scenarios/development-with-cloudflare/expected/first-deploy-with-demo-data.yaml
+++ b/tests/golden/scenarios/development-with-cloudflare/expected/first-deploy-with-demo-data.yaml
@@ -1644,7 +1644,7 @@ metadata:
     helm.sh/hook-delete-policy: before-hook-creation
 spec:
   backoffLimit: 0
-  activeDeadlineSeconds: 600
+  activeDeadlineSeconds: 1800
   template:
     spec:
       volumes:
diff --git a/tests/golden/scenarios/development-with-cloudflare/expected/first-deploy.yaml b/tests/golden/scenarios/development-with-cloudflare/expected/first-deploy.yaml
index 18b061c..eb1164d 100644
--- a/tests/golden/scenarios/development-with-cloudflare/expected/first-deploy.yaml
+++ b/tests/golden/scenarios/development-with-cloudflare/expected/first-deploy.yaml
@@ -1644,7 +1644,7 @@ metadata:
     helm.sh/hook-delete-policy: before-hook-creation
 spec:
   backoffLimit: 0
-  activeDeadlineSeconds: 600
+  activeDeadlineSeconds: 1800
   template:
     spec:
       volumes:
diff --git a/tests/golden/scenarios/escaping-env/expected/continuous.yaml b/tests/golden/scenarios/escaping-env/expected/continuous.yaml
index e6099fe..6d00e93 100644
--- a/tests/golden/scenarios/escaping-env/expected/continuous.yaml
+++ b/tests/golden/scenarios/escaping-env/expected/continuous.yaml
@@ -1842,7 +1842,7 @@ metadata:
     helm.sh/hook-delete-policy: before-hook-creation
 spec:
   backoffLimit: 0
-  activeDeadlineSeconds: 600
+  activeDeadlineSeconds: 1800
   template:
     spec:
       volumes:
diff --git a/tests/golden/scenarios/escaping-env/expected/first-deploy-with-demo-data.yaml b/tests/golden/scenarios/escaping-env/expected/first-deploy-with-demo-data.yaml
index 62dc8db..0fc45a9 100644
--- a/tests/golden/scenarios/escaping-env/expected/first-deploy-with-demo-data.yaml
+++ b/tests/golden/scenarios/escaping-env/expected/first-deploy-with-demo-data.yaml
@@ -1842,7 +1842,7 @@ metadata:
     helm.sh/hook-delete-policy: before-hook-creation
 spec:
   backoffLimit: 0
-  activeDeadlineSeconds: 600
+  activeDeadlineSeconds: 1800
   template:
     spec:
       volumes:
diff --git a/tests/golden/scenarios/escaping-env/expected/first-deploy.yaml b/tests/golden/scenarios/escaping-env/expected/first-deploy.yaml
index 765822f..777e939 100644
--- a/tests/golden/scenarios/escaping-env/expected/first-deploy.yaml
+++ b/tests/golden/scenarios/escaping-env/expected/first-deploy.yaml
@@ -1842,7 +1842,7 @@ metadata:
     helm.sh/hook-delete-policy: before-hook-creation
 spec:
   backoffLimit: 0
-  activeDeadlineSeconds: 600
+  activeDeadlineSeconds: 1800
   template:
     spec:
       volumes:
diff --git a/tests/golden/scenarios/production-with-cloudflare/expected/continuous.yaml b/tests/golden/scenarios/production-with-cloudflare/expected/continuous.yaml
index 00a720c..0ad3466 100644
--- a/tests/golden/scenarios/production-with-cloudflare/expected/continuous.yaml
+++ b/tests/golden/scenarios/production-with-cloudflare/expected/continuous.yaml
@@ -1945,7 +1945,7 @@ metadata:
     helm.sh/hook-delete-policy: before-hook-creation
 spec:
   backoffLimit: 0
-  activeDeadlineSeconds: 600
+  activeDeadlineSeconds: 1800
   template:
     spec:
       volumes:
diff --git a/tests/golden/scenarios/production-with-cloudflare/expected/first-deploy-with-demo-data.yaml b/tests/golden/scenarios/production-with-cloudflare/expected/first-deploy-with-demo-data.yaml
index 55fc7f9..c264a41 100644
--- a/tests/golden/scenarios/production-with-cloudflare/expected/first-deploy-with-demo-data.yaml
+++ b/tests/golden/scenarios/production-with-cloudflare/expected/first-deploy-with-demo-data.yaml
@@ -1945,7 +1945,7 @@ metadata:
     helm.sh/hook-delete-policy: before-hook-creation
 spec:
   backoffLimit: 0
-  activeDeadlineSeconds: 600
+  activeDeadlineSeconds: 1800
   template:
     spec:
       volumes:
diff --git a/tests/golden/scenarios/production-with-cloudflare/expected/first-deploy.yaml b/tests/golden/scenarios/production-with-cloudflare/expected/first-deploy.yaml
index 94be774..cca513b 100644
--- a/tests/golden/scenarios/production-with-cloudflare/expected/first-deploy.yaml
+++ b/tests/golden/scenarios/production-with-cloudflare/expected/first-deploy.yaml
@@ -1945,7 +1945,7 @@ metadata:
     helm.sh/hook-delete-policy: before-hook-creation
 spec:
   backoffLimit: 0
-  activeDeadlineSeconds: 600
+  activeDeadlineSeconds: 1800
   template:
     spec:
       volumes: