Unreliable zip deployment to the staging slot.

Hi there,

We are experiencing an intermediate issue when deploying our function app using Azure CLI as a zip package to the **staging** slot (See errors below).

Our function is configured to use:

1. Consumption plan.
2. Node.js 12
3. West US
4. Slots: staging, production

Additional tools:

azure-cli 2.16.0
core 2.16.0
telemetry 1.0.6

Your CLI is up-to-date.

### Here is the automated build process:

1. Update app settings in the **staging** slot

```
az webapp config appsettings set --slot staging --resource-group zzzResource --name zzz-test-us-dev --settings --verbose
```

2. Deploy to the **staging** slot

```
az functionapp deployment source config-zip --resource-group zzzResource --name zzz-test-us-dev --src --slot staging --verbose
```

3. Pinging **staging** slot instance to keep it warm for a minute.

4. Executing slot swap

```
az functionapp deployment slot swap --action swap --resource-group zzzResource --name zzz-test-us-dev --slot staging --target-slot production --verbose
```

### Additional app settings:

WEBSITE_ADD_SITENAME_BINDINGS_IN_APPHOST_CONFIG=1
WEBSITE_RUN_FROM_PACKAGE=1

This works most of the time, but occasionally, the deployment to the staging slot fails, and it ends up taking down the production slot with it.

### Errors:

```
Deploying "zzz-test-us-dev" to the "US" region.
Getting 'zzz-test-us-dev' app settings
Updating app version to 3
Done zipping the "zzz-test-us-dev" function!

Updating 'zzz-test-us-dev' app settings
Command ran in 1.837 seconds (init: 0.061, invoke: 1.776)

Deploying Azure function 'zzz-test-us-dev.'

Getting scm site credentials for zip deployment
Starting zip deployment. This operation can take a while to complete ...
Deployment endpoint responded with status code 202
Configuring default logging for the app, if not already enabled

Zip deployment failed. {'id': '6b96f759730146e09f4afb809c420b87', 'status': 3, 'status_text': '', 'author_email': 'N/A', 'author': 'N/A', 'deployer': 'ZipDeploy', 'message': 'Created via a push deployment', 'progress': '', 'received_time': '2020-12-14T20:08:34.2407885Z', 'start_time': '2020-12-14T20:08:34.2407885Z', 'end_time': '2020-12-14T20:08:34.8501901Z', 'last_success_end_time': None, 'complete': True, 'active': False, 'is_temp': False, 'is_readonly': True, 'url': 'https://zzz-test-us-dev-staging.scm.azurewebsites.net/api/deployments/latest', 'log_url': 'https://zzz-test-us-dev-staging.scm.azurewebsites.net/api/deployments/latest/log', 'site_name': 'zzz-test-us-dev', 'provisioningState': None}. Please run the command az webapp log deployment show -n zzz-test-us-dev -g zzzResource
Command ran in 17.114 seconds (init: 0.066, invoke: 17.048)

Unable to deploy the function app to the 'staging' slot. Error: Error: Command failed: az functionapp deployment source config-zip --resource-group zzzResource --name zzz-test-us-dev --src ./dist/zipPackages/zzz-test-us-dev.zip --slot staging --verbose

Getting scm site credentials for zip deployment
Starting zip deployment. This operation can take a while to complete ...
Deployment endpoint responded with status code 202
Configuring default logging for the app, if not already enabled

Zip deployment failed. {'id': '6b96f759730146e09f4afb809c420b87', 'status': 3, 'status_text': '', 'author_email': 'N/A', 'author': 'N/A', 'deployer': 'ZipDeploy', 'message': 'Created via a push deployment', 'progress': '', 'received_time': '2020-12-14T20:08:34.2407885Z', 'start_time': '2020-12-14T20:08:34.2407885Z', 'end_time': '2020-12-14T20:08:34.8501901Z', 'last_success_end_time': None, 'complete': True, 'active': False, 'is_temp': False, 'is_readonly': True, 'url': 'https://zzz-test-us-dev-staging.scm.azurewebsites.net/api/deployments/latest', 'log_url': 'https://zzz-test-us-dev-staging.scm.azurewebsites.net/api/deployments/latest/log', 'site_name': 'zzz-test-us-dev', 'provisioningState': None}. Please run the command az webapp log deployment show -n zzz-test-us-dev -g zzzResource
Command ran in 17.114 seconds (init: 0.066, invoke: 17.048)

```
Why would zip file deployment fail? The same zip file will get successfully deployed on a second or third attempt (IMHO: Not a zip package issue). In addition, there is no useful information in the error message.

We have seen this error quite a bit lately.

### **Production** instance ping results captured during the deployment to the **staging** slot:
```
>>> App Version <<< : 2 [Released On: 12/14/2020] - (76 ms)
>>> App Version <<< : 2 [Released On: 12/14/2020] - (80 ms)
>>> App Version <<< : 2 [Released On: 12/14/2020] - (77 ms)
>>> App Version <<< : 2 [Released On: 12/14/2020] - (89 ms)
>>> App Version <<< : 2 [Released On: 12/14/2020] - (96 ms)
>>> App Version <<< : 2 [Released On: 12/14/2020] - (81 ms)
>>> App Version <<< : 2 [Released On: 12/14/2020] - (450 ms)
>>> App Version <<< : 2 [Released On: 12/14/2020] - (102 ms)
>>> App Version <<< : 2 [Released On: 12/14/2020] - (356 ms)
>>> App Version <<< : 2 [Released On: 12/14/2020] - (85 ms)
>>> App Version <<< : 2 [Released On: 12/14/2020] - (107 ms)
>>> App Version <<< : 2 [Released On: 12/14/2020] - (69 ms)
>>> App Version <<< : 2 [Released On: 12/14/2020] - (85 ms)
>>> App Version <<< : 2 [Released On: 12/14/2020] - (77 ms)
ERROR: 503 - The service is unavailable. - (90 ms)
>>> App Version <<< : 2 [Released On: 12/14/2020] - (6043 ms)
>>> App Version <<< : 2 [Released On: 12/14/2020] - (104 ms)
>>> App Version <<< : 2 [Released On: 12/14/2020] - (74 ms)
>>> App Version <<< : 2 [Released On: 12/14/2020] - (92 ms)
>>> App Version <<< : 2 [Released On: 12/14/2020] - (106 ms)
>>> App Version <<< : 2 [Released On: 12/14/2020] - (73 ms)
>>> App Version <<< : 2 [Released On: 12/14/2020] - (70 ms)
>>> App Version <<< : 2 [Released On: 12/14/2020] - (98 ms)
>>> App Version <<< : 2 [Released On: 12/14/2020] - (97 ms)
>>> App Version <<< : 2 [Released On: 12/14/2020] - (100 ms)
>>> App Version <<< : 2 [Released On: 12/14/2020] - (89 ms)
>>> App Version <<< : 2 [Released On: 12/14/2020] - (91 ms)
>>> App Version <<< : 2 [Released On: 12/14/2020] - (85 ms)
>>> App Version <<< : 2 [Released On: 12/14/2020] - (89 ms)
```

How can we avoid failed deployment to **staging** slot bringing the **production** instance down (See the production instance ping results above with 503 error)?

Any help will be appreciated.