This repo contains scripts to build a CRAB 3 development testbed, assuming you're working on CERN lxplus system and interacting with the Openstack facility.
The typical workflow of the first time you use this script collection is:
git clone https://github.com/talamoig/crab-testbed # in an area that will be accessible by your VMs later on, eg. on afscd crab-testbed/cp paramsrc.sample paramsrc- customize
paramsrc(read below) ./vm-deploy.sh my-crab-server 'SLC6 CERN Server - x86_64 [2015-02-10]' # consider that if the vm already exists it will be reinstalled from scratch- wait until you can ssh to the VM (you can run
./vm-wait-then-ssh.sh my-crab-server) ssh my-crab-server(not if you logged using the previous script)cd path/to/crab-testbed./crabserver-install-part1.sh # at the end will reboot automaticallyssh my-crab-server./crabserver-install-part2.sh
During crabserver-install-part2.sh the configuration file paramsrc is used where shell-variable have to be defined. The paramaters are:
ORACLEUSER: your oracle userORACLEPASS: your oracle passwordGITUSER: your github user accountINITDB: if defined the oracle database will be initialized. Necessary only the first timeHGVER: version of the crab server. It is an optional parameter. If it is not configured, the last one will be used (you can obtain it withgit tag -l 'HG*'|tail -1in the git repositorygit://github.com/dmwm/deployment.git)REST_HOSTNAME: hostname of the RESTful CRAB3 interface, aka Crab ServerTW_HOSTNAME: hostname of the Task Worker. if not defined REST_HOSTNAME will be used insteadGISTEXTURL: (eg.https://gist.githubusercontent.com/talamoig/a46f05a991df431febb2/raw/gistfile1.txt) If you don't provide it, a suggestion of its content will be printed, according to the other parameters
Consider that you should have:
- a oracle account;
CRABServerandWMCorerepositories forked on your account.