wiki:OpenSpecimen HowTo Install

Version 19 (modified by Richard Bramley, 9 years ago) ( diff )

--

OpenSpecimen HowTo Install

Tags: OpenSpecimen HowTo

See https://openspecimen.atlassian.net/wiki/display/CAT/Deployment+steps

  • Have to make install.sh and catalina.sh executable.
  • Have to amend apache config:
    • Amend loadmodule.conf to:
      ...
      # Tomcat - used for JBoss
      #LoadModule jk_module /usr/lib64/apache2/mod_jk.so
      
      # Proxy to divert to the JBoss Application Server
      LoadModule proxy_module                   /usr/lib64/apache2-prefork/mod_proxy.so
      LoadModule proxy_http_module               /usr/lib64/apache2-prefork/mod_proxy_http.so
      
      
    • Amend httpd.cong to
      ...
      # Proxy to provide access to the JBoss Application Server
       ProxyPass / http://localhost:8280/ timeout=180
       ProxyPassReverse / http://localhost:8280/
       ProxyTimeout 180
      
      # Use mod_jk to forward to JBoss - not used now, cos it broke
      # Include /local/apache2/etc/mod.d/mod_jk.conf
      ...
      

Notes from watching Nitesh upgrade our dev version

  1. Our customisations are in the file /local/os_v20/app/plugin-dir/os-le-1.0.0.jar. This file can be copied between installations.
  2. He copied the build zip file (linked in the page above) into the directory /local/os_v20/installable and unzipped it.
  3. cd into the newly unzipped folder.
  4. `chmod 744 install.sh
  5. Run:
nohup ./install.sh -dhost localhost -dport 3306 -dname {db name} -duser {db user name} -dpass {db password name} -dtype mysql -ahome /local/{Open Specimen Directory}/app -ahost localhost -aport 8280 -adtype fresh -adatadir /local/{Open Specimen Directory}/app/os-data -aplugindir /local/{Open Specimen Directory}/app/plugin-dir &
  1. Run cp /local/{Open Specimen Directory}/installable/os-le-1.0.0.jar /local/{Open Specimen Directory}/app/plugin-dir
  1. Create the script run_open_specimen.sh:
#!/bin/bash

unset JAVA_TOOL_OPTIONS
export JAVA_ROOT=/usr/lib64/jvm/java
export JDK_HOME=/usr/lib64/jvm/java
export SDK_HOME=/usr/lib64/jvm/java
export JAVA_BINDIR=/usr/lib64/jvm/java/bin
export JRE_HOME=/usr/lib64/jvm/java/jre
JAVA_HOME=/opt/java
JRE_HOME=/opt/java/jre
OPEN_SPECIMEN_DIR=/local/{Open Specimen Directory}/app

export JAVA_TOOL_OPTIONS=-XX:CompressedClassSpaceSize=10m
export JAVA_OPTS="-Xmx2G -XX:-UseCompressedClassPointers"

$OPEN_SPECIMEN_DIR/bin/catalina.sh start >$OPEN_SPECIMEN_DIR/logs/tomcat-start.log  2>$OPEN_SPECIMEN_DIR/logs/tomcat-start.err &
  1. Create the script stop_open_specimen.sh:
#!/bin/bash

unset JAVA_TOOL_OPTIONS
export JAVA_ROOT=/usr/lib64/jvm/java
export JDK_HOME=/usr/lib64/jvm/java
export SDK_HOME=/usr/lib64/jvm/java
export JAVA_BINDIR=/usr/lib64/jvm/java/bin
export JRE_HOME=/usr/lib64/jvm/java/jre
JAVA_HOME=/opt/java
JRE_HOME=/opt/java/jre
OPEN_SPECIMEN_DIR=/local/openSpecimen_v2_4/app

export JAVA_TOOL_OPTIONS=-XX:CompressedClassSpaceSize=10m
export JAVA_OPTS="-Xmx2G -XX:-UseCompressedClassPointers"

$OPEN_SPECIMEN_DIR/bin/catalina.sh stop
  1. Give www-run account permission to run the scripts:
setfacl -m u:wwwrun:rx /local/{Open Specimen Directory}/run_open_specimen.sh
setfacl -m u:wwwrun:rx /local/{Open Specimen Directory}/stop_open_specimen.sh
  1. Give www-run permission to write to the logs and work directories:
setfacl -R -m u:wwwrun:rwx /local/{Open Specimen Directory}/app/logs/
setfacl -R -m u:wwwrun:rwx /local/{Open Specimen Directory}/app/work/
  1. Give www-run and yourself (go on, treat yourself) permission to run Tomcat:
    chmod 774 /local/{Open Specimen Directory}/app/bin/catalina.sh
    setfacl -m u:wwwrun:rx /local/{Open Specimen Directory}/app/bin/catalina.sh
    
  1. Then run_open_specimen.sh as yourself. You need to do this to be able to unpack the WAR file. The site will not come up successfully because there is a bug or somesuch in the database upgrade scripts.
  1. Nitesh changed one of the change sets because it fails if you don't. Hurray.
    1. The file is /local/{Open Specimen Directory}/app/webapps/openspecimen/WEB-INF/classes/db/2.0/schema.xml
    2. Change <changeSet author="vlonushte" id="FK on domain name of users table"> to <changeSet author="vlonushte" id="FK on domain name of users table" failOnError="false">

UPDATE WORKFLOW table to correct format

Run the command:

update os_cp_workflows set workflows = '[{"name":"registerParticipant","view":"plugin-ui-resources/le/collect-specimens.html","ctrl":"leRegAndCollectSpecimensCtrl","data":{"boxOpts":{"dimension":{"rows":8,"columns":12},"compact":false}}}]';

New Process

  1. Unzip Zip file
  2. Run install.sh
  3. sudo -u wwwrun ./first_run_open_specimen.sh
  4. Check for process still running ps -ef|grep cat
  5. If the process is still running, kill it using sudo -u wwwrun kill {process id}
  6. sudo -u wwwrun ./run_open_specimen.sh
  7. Check it starts by running tail -f app/logs/catalina.out

Error: Macro BackLinks(None) failed
'Environment' object has no attribute 'get_db_cnx'

Note: See TracWiki for help on using the wiki.