Azure-cli: az functionapp Azure Cli deploy error (v2)

[allschu]

This is autogenerated. Please review and update as needed.

Describe the bug

Command Name
az functionapp deployment source config-zip

Errors:

HTTPSConnectionPool(host='*************.azurewebsites.net', port=443): Read timed out. (read timeout=3)
Traceback (most recent call last):
python3.6/site-packages/urllib3/connectionpool.py, ln 387, in _make_request
    six.raise_from(e, None)
<string>, ln 2, in raise_from
python3.6/site-packages/urllib3/connectionpool.py, ln 383, in _make_request
    httplib_response = conn.getresponse()
lib/python3.6/http/client.py, ln 1331, in getresponse
    response.begin()
...
socket.timeout: The read operation timed out


python3.6/site-packages/requests/adapters.py, ln 529, in send
    raise ReadTimeout(e, request=request)
requests.exceptions.ReadTimeout: HTTPSConnectionPool(host='*****************.scm.azurewebsites.net', port=443): Read timed out. (read timeout=3)

To Reproduce:

Deploy zipped publish package to Azure Function App service. using the following command

az functionapp deployment source config-zip -g [RESOURCE-GROUP ]-n [ FUNCTION APP NAME] --src publish.zip

Expected Behavior

Environment Summary

Linux-4.15.0-1082-azure-x86_64-with-debian-stretch-sid
Python 3.6.5
Installer: DEB

azure-cli 2.5.1 *

Additional Context

[msftbot[bot]]

Thanks for the feedback! We are routing this to the appropriate team for follow-up. cc @ahmedelnably, @fabiocav.

[yonzhan]

functionapp

[ttjackott]

We get this problem as well, it can be quite intermittent

[lobsteropteryx]

We've seen this as well, intermittently across many projects. Have opened an Azure support issue, but we never got this resolved.

[katerd]

We're seeing exactly the same read timeout error too. One of our pipelines deploys 5 function apps in sequence, and it's a dice roll if it succeeds. We're considering retry loops for each of the az commands as a workaround.

[ttjackott]

We're seeing exactly the same read timeout error too. One of our pipelines deploys 5 function apps in sequence, and it's a dice roll if it succeeds. We're considering retry loops for each of the az commands as a workaround.

We do exactly that - we retry up to 3 times sometimes to work around this. Here is our workaround powershell which we do via an AzureCLI task in our yaml multi stage pipelines:

$resourceGroupName = "...."
$functionAppName = "...."
$artifactPath = "...."

function Deploy-Function {
    return az functionapp deployment source config-zip --resource-group $resourceGroupName --name $functionAppName --src $artifactPath
}

$functionDeploymentResult = $null;
For ($i = 0; $i -lt 3; $i++) {
    Write-Host "Function deployment attempt $($i+1)"

    # successful results are returned as JSON, if this converts to JSON we can assume it is successful
    # failed results do not parse as JSON, so will be null on failure
    $functionDeploymentResult = Deploy-Function | ConvertFrom-Json
    if ($functionDeploymentResult){
        break
    }
    else {
        Write-Host "Function deployment failed at attempt $($i+1)"
    }
}

[arthfl]

Is there any progress on this issue?
We have failing function deployments all over the place because of this.

[shuruev]

Having the same issue, started being worse recently. Not sure what it really depends on.

[arthfl]

We found btw. out that (at least for our simple use-case), the app deployment works perfectly fine when you use az webapp instead of az functionapp.
This makes this whole issue a bit weirder, but maybe this is helping somebody ;-)

[sonic1981]

I think I finally got to the bottom of this issue. It seems to be a resource problem, kudos to https://github.com/Azure/azure-cli/issues/11980#issuecomment-685823613.

I upgraded my service plan to a premium one and the problem seems to have gone away