This is the main deployment script.

./ -h

 DOCS:                 _____
_________________________________________________________________________/ v0.9

  Loads one or more profiles profile to deploy an application.
  If the download file is not newer then it does not extract files and does not
  Optionally it cleans up the target directory.
  Runs pre and post hooks - it updates the config files only and sets the owner.


  -h | show this help and exit
  -f | force full installation even if the download file is not newer
  -l | list exiting profile names

  Set one or more valid profile names. By default it loops over all profiles.
  This prameter limits the execution to the given profiles.
  Use option -l to get a list of profiles.

If you start it without parameter it will loop over all existing profiles. You can add an existing profile name to limit the execution to that profile only.

To start a single profile start ./ -l to get a list of exisring configured profiles and then ./ [PROFILE].

Each execution on CLI or cronjob creates a log file per profile in /var/log/imldeployment-client/.

> ls -ltr /var/log/imldeployment-client/
total 60
-rw-r--r-- 1 root root 28613 Apr 25 15:09 profile_A__2022-04-25__150405.log
-rw-r--r-- 1 root root 30747 Apr 25 15:10 profile_B__2022-04-25__150959.log

Each logfile contens the profile name + 2 underscore + timestamp. There is an automatic cleanup: the script keeps the last 10 logs per profile.


The script is used to read a template (*.erb) to replace simple placeholders by regex <\%\=\ *\@replace\[\"$key\"\]\ *\%> with a value coming from one or more replacement data files. The output will be written to a target file.

===== IML DEPLOYMENT :: replace variables in erb syntax =====

template    : 
output      : 
replacements: replace*.txt



With you can access the software archive.

  • read phase names
  • read ci project ids available in a given phase
  • list files of a given projects of a phase
  • download a single file to a custom target file
  • download ALL files to current working directory


  -c CFGFILE  load custom config file after defaults in
  -d          enable debug infos
  -e PHASE    phase; overrides env variable IMLCI_PHASE
  -f FILE     filename to get (without path); overrides env variable IMLCI_FILE
  -l ITEM     list
  -o OUTFILE  optional output file
  -p PROJECT  ci project id; overrides env variable IMLCI_PROJECT
  -s SECRET   override secret in IMLCI_PKG_SECRET
  -u URL      URL of iml ci server without trailing /; overrides env variable IMLCI_URL


  CFGFILE     custom config file. It is useful to handle files of different 
              projects on a server.
  PHASE       is a phase of the ci server; one of preview|stage|live
  FILE        is a filename without path that was created by ci server.
  OUTFILE     Output file. It can countain a path. If none is given the filename
              will be taken from FILE and stored in current directory
  PROJECT     project id of the ci server
  SECRET      secret to access project data on package server. Your given secret
              must match the secret on package server to get access to any url.
  ITEM        type what to list; one of phases|projects|files
              To list projects a phase must be set.
              To list files a phase and a project must be set.


  You don't need to set all values by command line. Use a config to set defaults


  If url, secret, project and phase are set in the config you can operate by
  setting the filename to request. -f FILE 
    downloads FILE to the current dir. -f FILE -o my-own-filename.tgz 
    downloads FILE as my-own-filename.tgz -f ALL 
    there is a special file ALL; it fetches all filenames by executing a directory 
    listing and then downloads all remote files with their original name -e preview -l projects
    list existing projects in phase preview -l files
    list existing files of current project

  Remark: The directory listing can be turned off on the package server and
  results in a 403 status.