Merge pull request #53 from netfoundry/release-v5.11.3

run ci test in aws

Author: qrkourier

.github/workflows/main.yml

@@ -72,33 +72,13 @@ jobs:
         register-python-argcomplete nfctl
 
     - name: Run the NF CLI demo to  test installed version
+      id: test_demo
       shell: bash
       env:
         NETFOUNDRY_CLIENT_ID: ${{ secrets.NETFOUNDRY_CLIENT_ID }}
         NETFOUNDRY_PASSWORD: ${{ secrets.NETFOUNDRY_PASSWORD }}
         NETFOUNDRY_OAUTH_URL: ${{ secrets.NETFOUNDRY_OAUTH_URL }}
-      run: |
-        set -o xtrace
-        set -o pipefail
-
-        nfctl config \
-          general.network=$(nfctl demo --echo-name --prefix 'gh-${{ github.run_id }}') \
-          general.yes=True \
-          general.verbose=yes || true # FIXME: sometimes config command exits with an error
-        nfctl demo \
-          --size medium \
-          --regions us-ashburn-1 us-phoenix-1 \
-          --provider OCI
-        nfctl \
-          list services
-        nfctl \
-          get service name=echo% > /tmp/echo.yml
-        nfctl \
-          delete service name=echo%
-        nfctl \
-          create service --file /tmp/echo.yml
-        nfctl \
-          delete network
+      run: ./scripts/test-demo.sh
 
     - name: Publish Test Package
       uses: pypa/gh-action-pypi-publish@v1.13.0
@@ -162,3 +142,43 @@ jobs:
         platforms: linux/amd64,linux/arm64
         push: true
         tags: ${{ steps.compose_tags.outputs.container_tags }}
+
+  cleanup-delay:
+    if: failure()
+    needs: [build_pypi_and_docker]
+    runs-on: ubuntu-latest
+    steps:
+      - name: Wait 30 minutes before cleanup
+        run: |
+          echo "Test demo failed to complete. Waiting 30 minutes before cleanup to allow investigation..."
+          sleep 1800
+
+  cleanup-network:
+    if: always() && needs.build_pypi_and_docker.result == 'failure'
+    needs: [cleanup-delay]
+    runs-on: ubuntu-latest
+    steps:
+      - uses: actions/checkout@v5
+
+      - name: Set up Python
+        uses: actions/setup-python@v6
+        with:
+          python-version: '3.12'
+
+      - name: Install nfctl
+        run: |
+          python -m pip install --upgrade pip
+          pip install .
+
+      - name: Delete test network
+        env:
+          NETFOUNDRY_CLIENT_ID: ${{ secrets.NETFOUNDRY_CLIENT_ID }}
+          NETFOUNDRY_PASSWORD: ${{ secrets.NETFOUNDRY_PASSWORD }}
+          NETFOUNDRY_OAUTH_URL: ${{ secrets.NETFOUNDRY_OAUTH_URL }}
+        run: |
+          # Use wildcard pattern to match network created by this run
+          NETWORK_PATTERN="gh-${GITHUB_RUN_ID}-%"
+          echo "Attempting to delete network matching: ${NETWORK_PATTERN}"
+
+          # Try to delete the network, ignore errors if it doesn't exist
+          nfctl delete network "name=${NETWORK_PATTERN}" --yes || echo "Network may not exist or already deleted"

netfoundry/ctl.py

@@ -13,6 +13,7 @@
 import signal
 import jwt
 import tempfile
+from concurrent.futures import ThreadPoolExecutor, as_completed
 from json import dumps as json_dumps
 from json import load as json_load
 from json import loads as json_loads
@@ -22,6 +23,7 @@
 from subprocess import CalledProcessError
 from sys import exit as sysexit
 from sys import stderr, stdin, stdout
+from threading import Lock
 from xml.sax.xmlreader import InputSource
 
 from jwt.exceptions import PyJWTError
@@ -44,7 +46,7 @@
 
 # import milc cli
 from milc import cli, questions  # noqa: E402
-# set milc options using new API
+# set milc options (requires milc >= 1.8.0)
 cli.milc_options(name='nfctl', author='NetFoundry', version=f'v{netfoundry_version}')
 # this creates the config subcommand
 from milc.subcommand import config  # noqa: F401,E402
@@ -94,7 +96,7 @@ def __call__(self, parser, namespace, values, option_string=None):
 @cli.argument('-B', '--borders', default=True, action='store_boolean', help='print cell borders in text tables')
 @cli.argument('-H', '--headers', default=True, action='store_boolean', help='print column headers in text tables')
 @cli.argument('-Y', '--yes', action='store_true', arg_only=True, help='answer yes to potentially-destructive operations')
-@cli.argument('-W', '--wait', help='seconds to wait for long-running processes to finish', default=900)
+@cli.argument('-W', '--wait', type=int, help='seconds to wait for long-running processes to finish', default=900)
 @cli.argument('--proxy', help=argparse.SUPPRESS)
 @cli.argument('--gateway', default="gateway", help=argparse.SUPPRESS)
 @cli.entrypoint('configure the CLI to manage a network')
@@ -961,40 +963,63 @@ def demo(cli):
         else:
             spinner.succeed(f"Found a hosted router in {region}")
 
-    spinner.text = f"Creating {len(fabric_placements)} hosted router(s)"
-    with spinner:
-        for region in fabric_placements:
-            er_name = f"Hosted Router {region} [{cli.config.demo.provider}]"
-            if not network.edge_router_exists(er_name):
-                er = network.create_edge_router(
-                    name=er_name,
-                    attributes=[
-                        "#hosted_routers",
-                        "#demo_exits",
-                        f"#{cli.config.demo.provider}",
-                    ],
-                    provider=cli.config.demo.provider,
-                    location_code=region,
-                    tunneler_enabled=False,  # workaround for MOP-18098 (missing tunneler binding in ziti-router config)
-                )
-                hosted_edge_routers.extend([er])
-                spinner.succeed(f"Created {cli.config.demo.provider} router in {region}")
+    # Helper function to create or validate a single router (runs in parallel)
+    def create_or_validate_router(region):
+        """Create or validate router for a region. Returns (region, router_dict, message)."""
+        er_name = f"Hosted Router {region} [{cli.config.demo.provider}]"
+        if not network.edge_router_exists(er_name):
+            er = network.create_edge_router(
+                name=er_name,
+                attributes=[
+                    "#hosted_routers",
+                    "#demo_exits",
+                    f"#{cli.config.demo.provider}",
+                ],
+                provider=cli.config.demo.provider,
+                location_code=region,
+                tunneler_enabled=False,  # workaround for MOP-18098 (missing tunneler binding in ziti-router config)
+            )
+            message = f"Created {cli.config.demo.provider} router in {region}"
+            return (region, er, message)
+        else:
+            er_matches = network.edge_routers(name=er_name, only_hosted=True)
+            if len(er_matches) == 1:
+                er = er_matches[0]
             else:
-                er_matches = network.edge_routers(name=er_name, only_hosted=True)
-                if len(er_matches) == 1:
-                    er = er_matches[0]
-                else:
-                    raise RuntimeError(f"unexpectedly found more than one matching router for name '{er_name}'")
-                if er['status'] in RESOURCES["edge-routers"].status_symbols["error"] + RESOURCES["edge-routers"].status_symbols["deleting"] + RESOURCES["edge-routers"].status_symbols["deleted"]:
-                    raise RuntimeError(f"hosted router '{er_name}' has unexpected status '{er['status']}'")
+                raise RuntimeError(f"unexpectedly found more than one matching router for name '{er_name}'")
+            if er['status'] in RESOURCES["edge-routers"].status_symbols["error"] + RESOURCES["edge-routers"].status_symbols["deleting"] + RESOURCES["edge-routers"].status_symbols["deleted"]:
+                raise RuntimeError(f"hosted router '{er_name}' has unexpected status '{er['status']}'")
+            return (region, er, None)  # No message for existing routers
+
+    # Parallelize router creation with thread-safe spinner updates
+    spinner.text = f"Creating {len(fabric_placements)} hosted router(s)"
+    spinner_lock = Lock()
+    new_routers = []
+
+    with ThreadPoolExecutor(max_workers=min(len(fabric_placements), 5)) as executor:
+        # Submit all router creation tasks
+        future_to_region = {executor.submit(create_or_validate_router, region): region for region in fabric_placements}
+
+        # Collect results as they complete
+        for future in as_completed(future_to_region):
+            region, er, message = future.result()
+            new_routers.append(er)
+
+            # Thread-safe spinner update for newly created routers
+            if message:
+                with spinner_lock:
+                    spinner.succeed(message)
+
+    # Add all new routers to the list
+    hosted_edge_routers.extend(new_routers)
 
     if not len(hosted_edge_routers) > 0:
         raise RuntimeError("unexpected problem with router placements, found zero hosted routers")
 
     spinner.text = f"Waiting for {len(hosted_edge_routers)} hosted router(s) to provision"
     with spinner:
         for router in hosted_edge_routers:
-            network.wait_for_statuses(expected_statuses=RESOURCES["edge-routers"].status_symbols["complete"], id=router['id'], type="edge-router", wait=2222, progress=False)
+            network.wait_for_statuses(expected_statuses=RESOURCES["edge-routers"].status_symbols["complete"], id=router['id'], type="edge-router", wait=cli.config.general.wait, progress=False)
             # ensure the router tunneler is available
             # network.wait_for_entity_name_exists(entity_name=router['name'], entity_type='endpoint')
             # router_tunneler = network.find_resources(type='endpoint', name=router['name'])[0]
@@ -1091,31 +1116,6 @@ def demo(cli):
                 services[svc]['properties'] = network.services(name=svc)[0]
                 spinner.succeed(sub("Finding", "Found", spinner.text))
 
-    # create a customer-hosted ER unless exists
-    customer_router_name = "Branch Exit Router"
-    spinner.text = f"Finding customer router '{customer_router_name}'"
-    with spinner:
-        if not network.edge_router_exists(name=customer_router_name):
-            spinner.text = sub("Finding", "Creating", spinner.text)
-            customer_router = network.create_edge_router(
-                name=customer_router_name,
-                attributes=["#branch_exit_routers"],
-                tunneler_enabled=True)
-        else:
-            customer_router = network.edge_routers(name=customer_router_name)[0]
-            spinner.succeed(sub("Finding", "Found", spinner.text))
-
-    spinner.text = f"Waiting for customer router {customer_router_name} to be ready for registration"
-    # wait for customer router to be PROVISIONED so that registration will be available
-    with spinner:
-        try:
-            network.wait_for_statuses(expected_statuses=RESOURCES["edge-routers"].status_symbols["complete"], id=customer_router['id'], type="edge-router", wait=222, progress=False)
-            customer_router_registration = network.rotate_edge_router_registration(id=customer_router['id'])
-        except Exception as e:
-            raise RuntimeError(f"error getting router registration, got {e}")
-        else:
-            spinner.succeed(f"Customer router ready to register with key '{customer_router_registration['registrationKey']}'")
-
     # create unless exists
     app_wan_name = "Default Service Policy"
     spinner.text = "Finding service policy"

netfoundry/network.py

@@ -458,6 +458,14 @@ def patch_resource(self, patch: dict, type: str = None, id: str = None, wait: in
                 headers=headers,
                 json=pruned_patch
             )
+            if after_response.status_code in range(400, 600):
+                self.logger.debug(
+                    '%s\n%s %s\r\n%s\r\n\r\n%s',
+                    '-----------RESPONSE-----------',
+                    after_response.status_code, after_response.reason,
+                    '\r\n'.join('{}: {}'.format(k, v) for k, v in after_response.headers.items()),
+                    after_response.text
+                )
             after_response.raise_for_status()  # raise any gross errors immediately
             after_response_code = after_response.status_code
             if after_response_code in [STATUS_CODES.codes.OK, STATUS_CODES.codes.ACCEPTED]: