Cannot start service elasticsearch: driver failed programming external connectivity on endpoint

I am facing the following error from time to time within a new Gitpod when running docker-compose up -d

docker-compose up -d
Creating network "stack_default" with the default driver
Creating network "stack_dockernet" with driver "bridge"
Creating stack_mysql_1 ... 
Creating stack_embed_1 ... 
Creating stack_filebeat_1 ... 
Creating stack_elasticsearch_1 ... 
Creating stack_mysql_1             ... done
Creating stack_embed_1             ... done
Creating stack_filebeat_1          ... done
Creating stack_redis_1             ... 
Creating stack_app_1               ... done
Creating stack_mongo_1             ... done
Creating stack_sync-core-1-mongo_1 ... done
Creating stack_redis_1             ... doneCreating stack_rabbitmq_1          ... done
Creating stack_drupal_1            ... done
Creating stack_sync-core-1_1       ... done

ERROR: for elasticsearch  Cannot start service elasticsearch: driver failed programming external connectivity on endpoint stack_elasticsearch_1 (5c874fc0de6095ebe3b3641b14dedfb24a219d48ecc154fdf0b368abb59f4b93): Error starting userland proxy: cannot expose slirp4net port: error reply: map[desc:bad request: add_hostfwd: slirp_add_hostfwd failed]
ERROR: Encountered errors while bringing up the project.

https://blue-rooster-cpbzo65k.ws-eu18.gitpod.io/

Hmm…are you able to share a minimal, reproducible example for us to investigate?

Hi,
no I have no clue how I could reproduce this since it only happens from time to time.
I just woke up the workspace from sleep and the error went away.
Is there any chance for a faster support when I am again facing the error so that I can show it to you right within the pod?

Thanks for the update, @sgurit! Please let us know if it happens again, it is difficult to help when we cannot reproduce it from our side as you can imagine.

Is there any chance for a faster support when I am again facing the error so that I can show it to you right within the pod?

Right now, we tend to reply within our working hours (i.e. 9AM - 5PM CET, Mon-Fri) Whereas, evenings and weekends may take us a little bit longer. Next time it does crop up, please bump this thread again and we’ll try and jump in :slight_smile: