Hibernate Search on Amazon Beanstalk

I am trying to create a hibernate search project and deploy it on amazon beanstalk.

The project works fine locally but i have the following doubts regarding Beanstalk

1. I am guessing beanstalk might use multiple underlying EC2 instances.If this is correct on which instance should i create the lucene index directory

2.If it is not possible to create a local filesystem setup is there a way i can use Amazon S3 as the index storage which will be common.