Get hands-on experience with 20+ free Google Cloud products and $300 in free credit for new customers.

GCP Batch In-use IP addresses

Hi team,

My "In-use IP addresses" limit is 64 and the Batch job quickly reaches that limit.  I want to run more tasks in parallel and utilize more machines.  I tried to increase this quota but was constantly rejected so I am wondering if  there is a way to reduce the "In-use IP addresses" usage and maintain high task parallelism? Thanks!

 

2 3 461
3 REPLIES 3

Do you need external IP addresses for your tasks?  If not, you can consider enabling noExternalIPAddres for your jobs. If you do need external IP addresses, you can try to pick larger machines so that more tasks can run per VM (one external IP per VM).

Thanks for the suggestion. My job does not need external ip address. However,  the api page https://cloud.google.com/batch/docs/reference/rest/v1/projects.locations.jobs#Job says that I need to do  additional configuration if I set  noExternalIPAddres to be True. I am wondering if there is anything else I need to do? Is there a tutorial on this?

Yes, there is some additional network configuration needed. Please start with Batch networking page, in particular, Block external access for a job.