How to install biowep (draft 0.1, May 2, 2006, by Paolo Romano) System prerequisites - Java (at least 1.4.1, currently in use 1.5.0)) See http://java.sun.com/ - Tomcat (5.5, currently in use 5.5.12) See http://tomcat.apache.org/ - MySQL (4, currently in use 4.1.12) See http://www.mysql.com/ The system has been developed under Windows, but it is currently only run on Linux. The production site is hosted on a Red Hat AS 4.0 machine. What it is included with biowep - Almost all Taverna Workbench 1.3.1 See http://taverna.sourceforge.net/ Pre-installation In order to run biowep on your machine, you need to have Tomcat and MySQL up and running. What to download biowep distribution includes the following five archives: - biowep core, including the user interface and all software that is needed to enact Scufl workflows, taken from the Taverna Workbench distribution - biowep database, including the definition of all tables that are needed by the user - biowep basic data, including a dump of the basic data that is needed to annotate workflows - biowep workflows, including workflows and related diagrams that are made available as a set of working examples (the annotation is not presently included) - biowep admin, including some old applications allowing for the insertion of workflows in the system (currently not maintained) In order to start working with biowep, you only need to download the core and the database. If you want to use our classification of processors, you must also download basic data. Workflows are there as a set of working examples, so you can take them for testing your site. Just adding them to the webapps tree is not enough for using them, you must also add them to the database. All archives are named according to the following general scheme: biowep-... where: - is the biowep version number, currently 1.0 - indicates the contents of the archive and can be one of the following names: core, database, basicdata, workflows, admin. - is the date when the file was created; it is in the form YYYYMMDD, e.g. Jan 15, 2006 --> 20060115 - is the final extension, specifying the file format; it presently can either be dump.gz (MySQL dumps compressed by gzip) or tar.gz (tar archives compressed by gzip) QUICK INSTALL We assume all prerequisites are satisfied! NB! you will see that all command lines end with an echo command. This is not needed at all... but if grep from this file all lines with echo in it, you will retrieve all commands that are needed to install biowep. Step 1: download biowep Retrieve the most recent version from the biowep web site: http://bioinformatics.hsanmartino.it/biowep/ . Follow instructions for downloading archives (after registration!). Download core, database and basic data. Step 2: copy biowep core into tomcat tree This step assumes that you have write rights in the tomcat webapps directory. This is usually only possible when you are logged in as root. First, create a new directory "biowep" mkdir biowep; echo 'done mkdir biowep' Then, uncompress (using gunzip and tar) the core archive in that directory: tar -xzvf biowep-1.0.core.20060424.tar.gz; echo 'done tar -xzvf biowep-1.0.core.20060424.tar.gz' You should now have a directory called biowep and a file called biowep.xml . Step 3: configure your portal Enter the newly created biowep directory. cd biowep; echo 'cd biowep' You should see a directory called conf. There, you will find a configuration file called conf.txt . You must edit this file in order to adapt the portal to your installation. Do not change the "agentconf.properties" file! This include a link to the Hermes server at the University of Camerino and MUST be left unchanged if you want that agent-based workflows created with BioWMS work. Just adapt the conf.txt parameters file. NB! If you don't adapt parameters to your site and machine, the portal won't work. cd conf; echo 'cd conf' The following parameters MUST be defined: serverAddress this is the main URL for your implementation, e.g.: serverAddress = "http://bioinformatics.hsanmartino.it:8090/biowep/"; Normally, you should just change the host name with yours. dbhost, dbuser and dbpass Quite self-explaining, the address (IP:port) od the server where your mysqld server is running, and the username and password for accessing biowep tables (you must have update rights) E.g.: dbhost = "10.1.7.23:3306"; dbuser = "biowep"; dbpass = "alcatraz2007"; smtp, from, nuser and pwdAuth Parameters for sending results by email when the workflow takes more than you are willing to wait for (see parameter time below). smtp is the IP of the SMTP server, from is the name of the sender as you want it to appear, nuser is a valid email address for the SMTP server and pwdAuth is teh password for that account. E.g.: smtp="10.1.7.15"; from="biowep@machine.domain"; nuser="biowep@machine.domain"; pwdAuth="singsing1999"; You don't have to change the remaining parameters. Anyway, here they are, together with their meanings: workflowRepository Name of the directory where workflows are stored. Default: workflowRepository = "workflow"; workflowImageRepository Name of the directory where diagrams of the workflows are stored. Default: workflowImageRepository = "workflow"; outputRepository Name of the directory where results of workflows' enactments are stored. Default: outputRepository = "output"; time Max time the system will wait for a workflow be executed, in seconds (1.000=1sec, 60.000=1min, 3.600.000=1hour) After this time, the system will return control to the user and the workflow will continue its operations in background. Results will be sent by email. Default is 2 mnutes: time = "120000"; giorni Time limit before removal of unsaved executions' results (days). Default: giorni="3";