Moving BBS back to FHCRC (for the time being)

Hutch machines:

• zin1 - linux, new release builds
• zin2 - linux, new devel builds
• moscato1 - windows (server 2008) new release builds
• moscato2 - windows (server 2008) new devel builds
• Start with BBS code. Make a new feature branch
• replace old machine names throughout with new (then go back and put in old machine names where they actually belong)
• make sure there are appropriate dirs under 3.2 and 3.3
• update README.md to reflect time zone, cloud>hutch changes...
• add info about getting into hutch network to credentials google doc
• make sure master builders have permission to push reports to S3
• update zins/moscatos, reboot if necessary ...

On hutch machines:

• remove svn's BBS working dir
• git clone BBS code
• remove old internal repos
• remove old bbs-X.Y-* dirs (done on zin2, not needed (?) on zin1)
• rsync internal repos from cloud machines
• install updated ensembl VEP script
• make sure chown-rootadmin is present in the right place
• set timezones back to west coast on macs
• update crontabs
• Make sure all machines can connect to master
• Install R-devel on zin2, moscato2
• install R-3.2.2 on zin1, R-patched on moscato1

in bioconductor.org

• update config.yaml with new builders
• Brian - add config helper scripts

in cloud

• Make AMIs of cloud machines (linux1/2, windows1/2)
• Terminate instances, remove elastic IPs