Deployment from scratch fails with latest changes in backend Lambda log group creation

[dlpzx]

Describe the bug

When deploying data.all in a fresh account using the latest code in main the CICD fails in the deployment of the backend stack. The failure comes from the creation of CloudWatch alarms for the log-groups of the backend lambdas.
In this PR #1134 the behavior on how the log groups for the Lambda functions changes, instead of letting Lambda create the groups automatically, it explicitly checks for the log group name and then performs the check. Unfortunately, the output of from_log_group_name is not None even when the log group does not exist.

       esproxy_loggroup = logs.LogGroup.from_log_group_name(
            self, 'esproxyloggroup', f'/aws/lambda/{resource_prefix}-{envname}-esproxy'
        )
....

        self.elasticsearch_proxy_handler = _lambda.DockerImageFunction(
            self,
            'ElasticSearchProxyHandler',
            function_name=f'{resource_prefix}-{envname}-esproxy',
            log_group=esproxy_loggroup
            if esproxy_loggroup is not None
            else logs.LogGroup(
                self, 'esproxyloggroup', log_group_name=f'/aws/lambda/{resource_prefix}-{envname}-esproxy'
            ),

How to Reproduce

Deploy data.all with the current changes including #1134

Expected behavior

Smooth behavior, log groups created and cloudwatch alarms created referencing the created log groups

Your project

No response

Screenshots

No response

OS

n/a

Python version

n/a

AWS data.all version

2.3+additional PRs

Additional context

No response

[dlpzx]

From testing we can conclude:

1. If we do not specify the log_group in the Lambda creation, it will automatically create a LogGroup for the Lambda using the name of the function, which is exactly what we create using logs.LogGroup
2. If we create the logs.LogGroup but it already exists, the whole stack fails creation
3. if we create the logs.LogGroup and it did not exist before, it succeeds

I think because of number 2, the check was added, however the check does not effectively check the existence of a previous stack.

I am inclined to revert the changes and go for the previous setup in which we did not define the log-group and use the automatically created ones (like in 1). But maybe there is something that I am missing @mourya-33