Beanstalk environemnt not updating the log groups on Cloudwatch

Scenario

Hi, I have few beanstalks environemnts running ".NET Core running on 64bit Amazon Linux 2/2.6.3" or ".NET 6 running on 64bit Amazon Linux 2023" Environments have different size, but all are Load Balanced and runs intances t3.micro or t3.small. Application is runing correctly and all features of Beanstalk seems to be fine, machine replacement, upload of application, etc.

The environemt are configured to send logs to Cloudwatch for 5 and 7 days, depending of the application. The log groups below were created automatically during creation of each environment.

1. Environemnet NAME/environment-health.log
2. Environemnet NAME/var/log/eb-engine.log
3. Environemnet NAME/var/log/eb-hooks.log
4. Environemnet NAME/var/log/nginx/access.log
5. Environemnet NAME/var/log/nginx/error.log
6. Environemnet NAME/var/log/web.stdout.log

I have double checked my EC2 Instace profile and Service Role

EC2 Instance Profile Policies

1. AmazonCloudWatchEvidentlyFullAccess
2. AmazonEC2RoleforSSM
3. AmazonS3FullAccess
4. AmazonSQSFullAccess
5. AWSElasticBeanstalkMulticontainerDocker
6. AWSElasticBeanstalkWebTier
7. AWSElasticBeanstalkWorkerTier
8. Cloudwatch_beanstalk
9. CloudWatchEventsFullAccess*

Service Role Policies

1. AmazonS3FullAccess
2. AWSElasticBeanstalkEnhancedHealth
3. AWSElasticBeanstalkManagedUpdatesCustomerRolePolicy
4. CloudWatchEventsFullAccess

The problem:

In my environments Cloudwatch is not creating one log file for each instance, sometimes it create 3 logs, for environemnts with 6 intances, sometimes create only 1. Even when it create the log file for a specific instance, the log suddenly stop to update. This is happening with all environemnts and there is no aparent reason as it ocurs with no standard.
I have tried to login on each instance and reatart cloudwatch agent, I have tried to change the period of retention of environment, but nothing has fixed the problem. Doe someone have suggestions for the issue?

Thanks!

Tarik