diff --git a/.github/workflows/test-docker.yml b/.github/workflows/test-docker.yml
deleted file mode 100644
index 22861c6..0000000
--- a/.github/workflows/test-docker.yml
+++ /dev/null
@@ -1,37 +0,0 @@
-name: Tests on Docker
-on:
-  push:
-    branches:
-      - main
-  pull_request:
-
-jobs:
-  container:
-    runs-on: ubuntu-latest
-    steps:
-      - name: Checkout
-        uses: actions/checkout@v4
-
-      - name: Set up Python
-        uses: actions/setup-python@v4
-        with:
-          python-version: 3.11
-          cache: 'pip'
-
-      - name: Install dependencies
-        run: |
-          pip install -r requirements.txt
-          pip install -r requirements-test.txt
-
-      - name: Pull example spider
-        run: docker pull ghcr.io/q-m/scrapyd-k8s-spider-example
-
-      - name: Run scrapyd-k8s
-        run: |
-          cp scrapyd_k8s.sample-docker.conf scrapyd_k8s.conf
-          python -m scrapyd_k8s &
-          while ! nc -q 1 localhost 6800 </dev/null; do sleep 1; done
-          curl http://localhost:6800/daemonstatus.json
-
-      - name: Run tests
-        run: pytest -vv test_api.py
diff --git a/.github/workflows/test-k8s.yml b/.github/workflows/test-k8s.yml
deleted file mode 100644
index 63c9a64..0000000
--- a/.github/workflows/test-k8s.yml
+++ /dev/null
@@ -1,47 +0,0 @@
-name: Tests on Kubernetes
-on:
-  push:
-    branches:
-      - main
-  pull_request:
-
-jobs:
-  container:
-    runs-on: ubuntu-latest
-    steps:
-      - name: Checkout
-        uses: actions/checkout@v4
-
-      - name: Set up Python
-        uses: actions/setup-python@v4
-        with:
-          python-version: 3.11
-          cache: 'pip'
-
-      - name: Install dependencies
-        run: |
-          pip install -r requirements.txt
-          pip install -r requirements-test.txt
-
-      - name: Start minikube
-        uses: medyagh/setup-minikube@master
-
-      - name: Prepare Kubernetes environment
-        run: |
-          kubectl create secret generic example-env-secret --from-literal=FOO_1=bar
-          kubectl create configmap example-env-configmap --from-literal=FOO_2=baz
-          # already pull image so we don't have to wait for it later
-          minikube image pull ghcr.io/q-m/scrapyd-k8s-spider-example:latest
-
-      - name: Run scrapyd-k8s
-        run: |
-          cp scrapyd_k8s.sample-k8s.conf scrapyd_k8s.conf
-          python -m scrapyd_k8s &
-          while ! nc -q 1 localhost 6800 </dev/null; do sleep 1; done
-          curl http://localhost:6800/daemonstatus.json
-
-      - name: Run tests
-        run: |
-          TEST_MAX_WAIT=60 \
-          TEST_AVAILABLE_VERSIONS=latest,`skopeo list-tags docker://ghcr.io/q-m/scrapyd-k8s-spider-example | jq -r '.Tags | map(select(. != "latest" and (startswith("sha-") | not))) | join(",")'` \
-          pytest -vv --color=yes test_api.py
diff --git a/.github/workflows/test-manifest.yml b/.github/workflows/test-manifest.yml
deleted file mode 100644
index b6bf8f3..0000000
--- a/.github/workflows/test-manifest.yml
+++ /dev/null
@@ -1,61 +0,0 @@
-name: Test Kubernetes manifest
-on:
-  push:
-    branches:
-      - main
-  pull_request:
-
-jobs:
-  container:
-    runs-on: ubuntu-latest
-    steps:
-      - name: Checkout
-        uses: actions/checkout@v4
-
-      - name: Set up Python
-        uses: actions/setup-python@v4
-        with:
-          python-version: 3.11
-          cache: 'pip'
-
-      - name: Install dependencies
-        run: |
-          pip install -r requirements.txt
-          pip install -r requirements-test.txt
-
-      - name: Set up Docker Buildx
-        uses: docker/setup-buildx-action@v3
-
-      - name: Build container
-        uses: docker/build-push-action@v5
-        with:
-          context: .
-          push: false
-          load: true
-          tags: test:latest
-          cache-from: type=gha
-          cache-to: type=gha,mode=max
-
-      - name: Start minikube
-        uses: medyagh/setup-minikube@master
-
-      - name: Deploy to minikube
-        run: |
-          minikube image load test:latest
-          # already pull image so we don't have to wait for it later
-          minikube image pull ghcr.io/q-m/scrapyd-k8s-spider-example:latest
-          # load manifest
-          sed -i 's/\(imagePullPolicy:\s*\)\w\+/\1Never/' kubernetes.yaml
-          sed -i 's/\(image:\s*\)ghcr\.io\/q-m\/scrapyd-k8s:/\1test:/' kubernetes.yaml
-          sed -i 's/\(type:\s*\)ClusterIP/\1NodePort/' kubernetes.yaml
-          kubectl create -f kubernetes.yaml
-          # and wait for scrapyd-k8s to become ready
-          kubectl wait --for=condition=Available deploy/scrapyd-k8s --timeout=60s
-          curl --retry 10 --retry-delay 2 --retry-all-errors `minikube service scrapyd-k8s --url`/daemonstatus.json
-
-      - name: Run tests
-        run: |
-          TEST_BASE_URL=`minikube service scrapyd-k8s --url` \
-          TEST_MAX_WAIT=60 \
-          TEST_AVAILABLE_VERSIONS=latest,`skopeo list-tags docker://ghcr.io/q-m/scrapyd-k8s-spider-example | jq -r '.Tags | map(select(. != "latest" and (startswith("sha-") | not))) | join(",")'` \
-          pytest -vv --color=yes test_api.py
diff --git a/.github/workflows/test.yml b/.github/workflows/test.yml
new file mode 100644
index 0000000..373e1ec
--- /dev/null
+++ b/.github/workflows/test.yml
@@ -0,0 +1,152 @@
+name: Scapyd-k8s CI
+on:
+  push:
+    branches:
+      - main
+  pull_request:
+
+jobs:
+  test-unit:
+    runs-on: ubuntu-latest
+    steps:
+      - name: Checkout
+        uses: actions/checkout@v4
+
+      - name: Set up Python
+        uses: actions/setup-python@v4
+        with:
+          python-version: 3.11
+          cache: 'pip'
+
+      - name: Install dependencies
+        run: |
+          pip install -r requirements.txt
+          pip install -r requirements-test.txt
+      
+      - name: Run tests
+        run: pytest -vv --color=yes scrapyd_k8s/tests/unit/
+
+  test-docker:
+    runs-on: ubuntu-latest
+    steps:
+      - name: Checkout
+        uses: actions/checkout@v4
+
+      - name: Set up Python
+        uses: actions/setup-python@v4
+        with:
+          python-version: 3.11
+          cache: 'pip'
+
+      - name: Install dependencies
+        run: |
+          pip install -r requirements.txt
+          pip install -r requirements-test.txt
+
+      - name: Pull example spider
+        run: docker pull ghcr.io/q-m/scrapyd-k8s-spider-example
+
+      - name: Run scrapyd-k8s
+        run: |
+          cp scrapyd_k8s.sample-docker.conf scrapyd_k8s.conf
+          python -m scrapyd_k8s &
+          while ! nc -q 1 localhost 6800 </dev/null; do sleep 1; done
+          curl http://localhost:6800/daemonstatus.json
+
+      - name: Run tests
+        run: pytest -vv --color=yes scrapyd_k8s/tests/integration/
+
+  test-manifest:
+    container:
+    runs-on: ubuntu-latest
+    steps:
+      - name: Checkout
+        uses: actions/checkout@v4
+
+      - name: Set up Python
+        uses: actions/setup-python@v4
+        with:
+          python-version: 3.11
+          cache: 'pip'
+
+      - name: Install dependencies
+        run: |
+          pip install -r requirements.txt
+          pip install -r requirements-test.txt
+
+      - name: Set up Docker Buildx
+        uses: docker/setup-buildx-action@v3
+
+      - name: Build container
+        uses: docker/build-push-action@v5
+        with:
+          context: .
+          push: false
+          load: true
+          tags: test:latest
+          cache-from: type=gha
+          cache-to: type=gha,mode=max
+
+      - name: Start minikube
+        uses: medyagh/setup-minikube@master
+
+      - name: Deploy to minikube
+        run: |
+          minikube image load test:latest
+          # already pull image so we don't have to wait for it later
+          minikube image pull ghcr.io/q-m/scrapyd-k8s-spider-example:latest
+          # load manifest
+          sed -i 's/\(imagePullPolicy:\s*\)\w\+/\1Never/' kubernetes.yaml
+          sed -i 's/\(image:\s*\)ghcr\.io\/q-m\/scrapyd-k8s:/\1test:/' kubernetes.yaml
+          sed -i 's/\(type:\s*\)ClusterIP/\1NodePort/' kubernetes.yaml
+          kubectl create -f kubernetes.yaml
+          # and wait for scrapyd-k8s to become ready
+          kubectl wait --for=condition=Available deploy/scrapyd-k8s --timeout=60s
+          curl --retry 10 --retry-delay 2 --retry-all-errors `minikube service scrapyd-k8s --url`/daemonstatus.json
+
+      - name: Run tests
+        run: |
+          TEST_BASE_URL=`minikube service scrapyd-k8s --url` \
+          TEST_MAX_WAIT=60 \
+          TEST_AVAILABLE_VERSIONS=latest,`skopeo list-tags docker://ghcr.io/q-m/scrapyd-k8s-spider-example | jq -r '.Tags | map(select(. != "latest" and (startswith("sha-") | not))) | join(",")'` \
+          pytest -vv --color=yes scrapyd_k8s/tests/integration/
+  test-k8s:
+    container:
+    runs-on: ubuntu-latest
+    steps:
+      - name: Checkout
+        uses: actions/checkout@v4
+
+      - name: Set up Python
+        uses: actions/setup-python@v4
+        with:
+          python-version: 3.11
+          cache: 'pip'
+
+      - name: Install dependencies
+        run: |
+          pip install -r requirements.txt
+          pip install -r requirements-test.txt
+
+      - name: Start minikube
+        uses: medyagh/setup-minikube@master
+
+      - name: Prepare Kubernetes environment
+        run: |
+          kubectl create secret generic example-env-secret --from-literal=FOO_1=bar
+          kubectl create configmap example-env-configmap --from-literal=FOO_2=baz
+          # already pull image so we don't have to wait for it later
+          minikube image pull ghcr.io/q-m/scrapyd-k8s-spider-example:latest
+
+      - name: Run scrapyd-k8s
+        run: |
+          cp scrapyd_k8s.sample-k8s.conf scrapyd_k8s.conf
+          python -m scrapyd_k8s &
+          while ! nc -q 1 localhost 6800 </dev/null; do sleep 1; done
+          curl http://localhost:6800/daemonstatus.json
+
+      - name: Run tests
+        run: |
+          TEST_MAX_WAIT=60 \
+          TEST_AVAILABLE_VERSIONS=latest,`skopeo list-tags docker://ghcr.io/q-m/scrapyd-k8s-spider-example | jq -r '.Tags | map(select(. != "latest" and (startswith("sha-") | not))) | join(",")'` \
+          pytest -vv --color=yes scrapyd_k8s/tests/integration/
diff --git a/README.md b/README.md
index 2e35266..cf84b54 100644
--- a/README.md
+++ b/README.md
@@ -151,7 +151,7 @@ curl http://localhost:6800/listjobs.json
 {
   "finished":[],
   "pending":[],
-  "running":[{"id":"e9b81fccbec211eeb3b109f30f136c01","project":"example","spider":"quotes","state":"running"}],
+  "running":[{"id":"e9b81fccbec211eeb3b109f30f136c01","project":"example","spider":"quotes","state":"running", "start_time": "2012-09-12 10:14:03.594664"}],
   "status":"ok"
 }
 ```
diff --git a/scrapyd_k8s/__init__.py b/scrapyd_k8s/__init__.py
new file mode 100644
index 0000000..e69de29
diff --git a/scrapyd_k8s/launcher/docker.py b/scrapyd_k8s/launcher/docker.py
index 65a7375..19b3ffe 100644
--- a/scrapyd_k8s/launcher/docker.py
+++ b/scrapyd_k8s/launcher/docker.py
@@ -1,9 +1,9 @@
 import re
-import os
 import socket
 
 import docker
-from ..utils import native_stringify_dict
+
+from ..utils import format_iso_date_string, native_stringify_dict
 
 class Docker:
 
@@ -66,21 +66,29 @@ def cancel(self, project_id, job_id, signal):
             c.kill(signal='SIG' + signal)
         return prevstate
 
-    def _parse_job(self, c):
-        return {
-            'id': c.labels.get(self.LABEL_JOB_ID),
-            'state': self._docker_to_scrapyd_status(c.status),
-            'project': c.labels.get(self.LABEL_PROJECT),
-            'spider': c.labels.get(self.LABEL_SPIDER)
+    def _parse_job(self, container):
+        state = self._docker_to_scrapyd_status(container.status)
+        job = {
+            'id': container.labels.get(self.LABEL_JOB_ID),
+            'state': state,
+            'project': container.labels.get(self.LABEL_PROJECT),
+            'spider': container.labels.get(self.LABEL_SPIDER),
         }
 
-    def _get_container(self, project_id, job_id):
-        filters = { 'label': self.LABEL_JOB_ID + '=' + job_id }
+        if state in ['running', 'finished']:
+            job['start_time'] = format_iso_date_string(container.attrs['State']['StartedAt'])
+        if state == 'finished':
+            job['end_time'] = format_iso_date_string(container.attrs['State']['FinishedAt'])
+
+        return job
+
+    def _get_container(self, project_id: str, job_id: str):
+        filters = {'label': self.LABEL_JOB_ID + '=' + job_id}
         c = self._docker.containers.list(all=True, filters=filters)
         if not c:
             return None
-        c = c[0]
 
+        c = c[0]
         if c.labels.get(self.LABEL_PROJECT) != project_id:
             return None
 
diff --git a/scrapyd_k8s/launcher/k8s.py b/scrapyd_k8s/launcher/k8s.py
index 3c969b7..8669c64 100644
--- a/scrapyd_k8s/launcher/k8s.py
+++ b/scrapyd_k8s/launcher/k8s.py
@@ -1,11 +1,10 @@
 import os
+from signal import Signals
 
 import kubernetes
 import kubernetes.stream
-from signal import Signals
-from subprocess import check_output, CalledProcessError
 
-from ..utils import native_stringify_dict
+from ..utils import format_datetime_object, native_stringify_dict
 
 class K8s:
 
@@ -122,13 +121,24 @@ def cancel(self, project, job_id, signal):
         return prevstate
 
     def _parse_job(self, job):
-        return {
+        state = self._k8s_job_to_scrapyd_status(job)
+        _job = {
             'id': job.metadata.labels.get(self.LABEL_JOB_ID),
-            'state': self._k8s_job_to_scrapyd_status(job),
+            'state': state,
             'project': job.metadata.labels.get(self.LABEL_PROJECT),
-            'spider': job.metadata.labels.get(self.LABEL_SPIDER)
+            'spider': job.metadata.labels.get(self.LABEL_SPIDER),
         }
 
+        if state in ['running', 'finished']:
+            _job['start_time'] = format_datetime_object(job.status.start_time)
+        if state == 'finished':
+            # Finished is mapped to succeeded and pending. Only when a job has succeeded will it have a 
+            # completion time.
+            completion_time = job.status.completion_time
+            _job['end_time'] = format_datetime_object(completion_time) if completion_time else None
+
+        return _job
+
     def _get_job(self, project, job_id):
         label = self.LABEL_JOB_ID + '=' + job_id
         r = self._k8s_batch.list_namespaced_job(namespace=self._namespace, label_selector=label)
@@ -170,13 +180,12 @@ def _k8s_job_name(self, project, job_id):
 
     def _k8s_kill(self, pod_name, signal):
         # exec needs stream, which modifies client, so use separate instance
-        k8s = kubernetes.client.CoreV1Api()
         resp = kubernetes.stream.stream(
-            k8s.connect_get_namespaced_pod_exec,
+            self._k8s.connect_get_namespaced_pod_exec,
             pod_name,
             namespace=self._namespace,
             # this is a bit blunt, bit it works and is usually available
             command=['/usr/sbin/killall5', '-' + str(signal)],
             stderr=True
         )
-        # TODO figure out how to get return value
\ No newline at end of file
+        # TODO figure out how to get return value
diff --git a/scrapyd_k8s/tests/__init__.py b/scrapyd_k8s/tests/__init__.py
new file mode 100644
index 0000000..e69de29
diff --git a/scrapyd_k8s/tests/integration/__init__.py b/scrapyd_k8s/tests/integration/__init__.py
new file mode 100644
index 0000000..e69de29
diff --git a/test_api.py b/scrapyd_k8s/tests/integration/test_api.py
similarity index 90%
rename from test_api.py
rename to scrapyd_k8s/tests/integration/test_api.py
index 54939b7..c459379 100644
--- a/test_api.py
+++ b/scrapyd_k8s/tests/integration/test_api.py
@@ -1,9 +1,12 @@
 #!/usr/bin/env python3
+from datetime import datetime
 import os
-import pytest
 import requests
 import time
 
+from scrapyd_k8s.api import launcher
+from scrapyd_k8s.launcher.k8s import K8s
+from scrapyd_k8s.utils import TIME_FORMAT
 
 BASE_URL = os.getenv('TEST_BASE_URL', 'http://localhost:6800')
 AVAIL_PROJECTS = os.getenv('TEST_AVAILABLE_PROJECTS', 'example').split(',')
@@ -15,6 +18,7 @@
 MAX_WAIT = int(os.getenv('TEST_MAX_WAIT', '6'))
 STATIC_SLEEP = float(os.getenv('TEST_STATIC_SLEEP', '2'))
 
+
 def test_root_ok():
     response = requests.get(BASE_URL)
     assert response.status_code == 200
@@ -159,6 +163,16 @@ def test_scenario_cancel_running_finished_ok():
     # wait until the job has stopped
     listjobs_wait(jobid, 'finished')
     jobinfo = assert_listjobs(finished=jobid)
+    start_time, end_time = jobinfo.pop('start_time'), jobinfo.pop('end_time')
+    assert start_time and datetime.strptime(start_time, TIME_FORMAT)
+
+    # Only when a k8s job has succeeded will it have a completion time. Here we cancel
+    # the job and thus we expect the end_time to undefined.
+    if isinstance(launcher, K8s):
+        assert not end_time
+    else:
+        assert end_time and datetime.strptime(end_time, TIME_FORMAT)
+
     assert jobinfo == { 'id': jobid, 'project': RUN_PROJECT, 'spider': RUN_SPIDER, 'state': 'finished' }
     # then cancel it again, though nothing would happen
     response = requests.post(BASE_URL + '/cancel.json', data={ 'project': RUN_PROJECT, 'job': jobid })
@@ -178,11 +192,16 @@ def scenario_regular(schedule_args):
     # wait until the job is running
     listjobs_wait(jobid, 'running')
     jobinfo = assert_listjobs(running=jobid)
+    start_time = jobinfo.pop('start_time')
+    assert start_time and datetime.strptime(start_time, TIME_FORMAT)
     assert jobinfo == { 'id': jobid, 'project': RUN_PROJECT, 'spider': RUN_SPIDER, 'state': 'running' }
     # wait until the job has finished
     listjobs_wait(jobid, 'finished')
     # check listjobs output
     jobinfo = assert_listjobs(finished=jobid)
+    start_time, end_time = jobinfo.pop('start_time'), jobinfo.pop('end_time')
+    assert start_time and datetime.strptime(start_time, TIME_FORMAT)
+    assert end_time and datetime.strptime(end_time, TIME_FORMAT)
     assert jobinfo == { 'id': jobid, 'project': RUN_PROJECT, 'spider': RUN_SPIDER, 'state': 'finished' }
 
 def assert_response_ok(response):
diff --git a/scrapyd_k8s/tests/unit/__init__.py b/scrapyd_k8s/tests/unit/__init__.py
new file mode 100644
index 0000000..e69de29
diff --git a/scrapyd_k8s/tests/unit/test_utils.py b/scrapyd_k8s/tests/unit/test_utils.py
new file mode 100644
index 0000000..6305ae0
--- /dev/null
+++ b/scrapyd_k8s/tests/unit/test_utils.py
@@ -0,0 +1,18 @@
+from datetime import datetime
+import pytest
+from scrapyd_k8s.utils import format_datetime_object, format_iso_date_string
+
+
+def test_format_iso_date_string():
+    input_date = "2024-08-30T13:45:30.123456"
+    expected = "2024-08-30 13:45:30.123456"
+    assert format_iso_date_string(input_date) == expected
+
+def test_format_iso_date_string_invalid():
+    with pytest.raises(ValueError):
+        format_iso_date_string("invalid_date")
+
+def test_format_datetime_object():
+    input_datetime = datetime(2024, 8, 30, 13, 45, 30, 123456)
+    expected = "2024-08-30 13:45:30.123456"
+    assert format_datetime_object(input_datetime) == expected
diff --git a/scrapyd_k8s/utils.py b/scrapyd_k8s/utils.py
index 7d9c752..495c95a 100644
--- a/scrapyd_k8s/utils.py
+++ b/scrapyd_k8s/utils.py
@@ -1,3 +1,8 @@
+from datetime import datetime
+
+
+TIME_FORMAT = '%Y-%m-%d %H:%M:%S.%f'
+
 def _to_native_str(text, encoding="utf-8", errors="strict"):
     if isinstance(text, str):
         return text
@@ -25,4 +30,10 @@ def native_stringify_dict(dct_or_tuples, encoding="utf-8", keys_only=True):
             else:
                 v = _to_native_str(v, encoding)
         d[k] = v
-    return d
\ No newline at end of file
+    return d
+
+def format_iso_date_string(date_string):
+    return datetime.fromisoformat(date_string).strftime(TIME_FORMAT)
+
+def format_datetime_object(datetime_obj):
+    return datetime_obj.strftime(TIME_FORMAT)