How to set up / access the Spark UI for AWS Glue jobs and notebooks

You need to deploy you own history server to visualize those longs, you can use a CloudFormation template, a docker container (you can also run locally a history server from a Spark distribution but adding s3 support requires a bit of work). See steps here: https://docs.aws.amazon.com/glue/latest/dg/monitor-spark-ui-history.html