uima-ducc-duccdocs/src/site/tex/duccbook/part4/install.tex

% 
% Licensed to the Apache Software Foundation (ASF) under one
% or more contributor license agreements.  See the NOTICE file
% distributed with this work for additional information
% regarding copyright ownership.  The ASF licenses this file
% to you under the Apache License, Version 2.0 (the
% "License"); you may not use this file except in compliance
% with the License.  You may obtain a copy of the License at
% 
%   http://www.apache.org/licenses/LICENSE-2.0
% 
% Unless required by applicable law or agreed to in writing,
% software distributed under the License is distributed on an
% "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY
% KIND, either express or implied.  See the License for the
% specific language governing permissions and limitations
% under the License.
% 
\section{Overview}

DUCC is a multi-user, multi-system distributed application.
For first-time users a staged installation/verification methodology is recommended,
roughly as follows:

\begin{itemize}
    \item Single system installation - single node - all work runs with the credentials of the installer.
      
    \item Optionally add more nodes. 
      
    \item Enable multiple-user support - processes run with the credentials of the submitting user,
      while DUCC runs as user {\em ducc}.
      This step requires root authority on one or more machines.
      
    \item Enable CGroup containers. This step requires root authority on every DUCC machine.
\end{itemize}

When upgrading from an existing installation the {\em ducc\_update} script may be used
to replace the system files while leaving the site-specific configuration files in place. 
For more information see
\ifdefined\DUCCSTANDALONE
{\em ``ducc\_update''} in the Administrative Commands section of the DuccBook. 
\else
\hyperref[subsec:admin.ducc-update] {\em ducc\_update}. 
\fi

Since with this release the persistence data about completed work is stored in a database, 
additional upgrade steps are required to convert the older file-based data in order to preserve information 
about past work. 
For more information see
\ifdefined\DUCCSTANDALONE
{\em ``db\_create''} and {\em ``db\_loader''} in the Administrative Commands section of the DuccBook. 
\else
\hyperref[subsec:admin.db-create] {\em db\_create} and \hyperref[subsec:admin.db-loader] {\em db\_loader}. 
\fi


DUCC is distributed as a compressed tar file.  If building from source, this file will be created in your svn
trunk/target directory. The distribution file is in the form
\begin{verbatim}
   uima-ducc-[version]-bin.tar.gz
\end{verbatim}
where [version] is the DUCC version; for example, {\em uima-ducc-2.1.0-bin.tar.gz}.  This document will refer to the distribution
file as the ``$<$distribution.file$>$''.

\section{Software Prerequisites}
\label{sec:install.prerequisites}

Single system installation:

\begin{itemize}
  \item Reasonably current Linux.  DUCC has been tested on SLES 11, RHEL 6 \& 7, and Ubuntu 14.04 
    with 64-bit Intel, IBM Power (Big and Little Endian) hardware.
    
    {\em Note:} On some systems the default {\em user limits}
    for max user processes (ulimit -u) and nfiles (ulimit -n) are defined too
    low for DUCC. The shell login profile for user {\em ducc} should set the
    soft limit for max user processes to be the same as the hard limit
    (ulimit -u `ulimit -Hu`), and
    the nfiles limit raised above 1024 to at least twice the number of user
    processes running on the cluster.

  \item Python 2.x, where 'x' is 4 or greater.  DUCC has not been tested on Python 3.x.
  \item Java 7 or 8. DUCC has been tested and run using IBM and Oracle JDK 1.7 \& 1.8.
  \item Passwordless ssh for the userid running DUCC
\end{itemize}
  
Additional requirements for multiple system installation:

\begin{itemize}
  \item All systems must have a shared filesystem (such as NFS or GPFS) and common user space.
        The \$DUCC\_HOME directory must be located on a shared filesystem.
\end{itemize}
  
Additional requirements for running multiple user processes with their own credentials.

\begin{itemize}
  \item A userid {\em ducc}, and group {\em ducc}.  User {\em ducc} must the the only member of group {\em ducc}.
  \item DUCC run with user {\em ducc} credentials.
  \item Root access is required to setuid-root the DUCC process launcher.
\end{itemize}
  
Additional requirements for CGroup containers:

\begin{itemize}
  \item A userid {\em ducc}, and group {\em ducc}.  User {\em ducc} must the the only member of group {\em ducc}.
  \item DUCC run with user {\em ducc} credentials.
  \item libcgroup1-0.37+ on SLES, libcgroup-0.37+ on RHEL, and on Ubuntu all of cgroup-bin cgroup-lite libcgroup1
  \item along with a customized /etc/cgconfig.conf
\end{itemize}

  
In order to build DUCC from source the following software is also required:
\begin{itemize}
    \item A Subversion client, from \url{http://subversion.apache.org/packages.html}.  The
      svn url is \url{https://svn.apache.org/repos/asf/uima/uima-ducc/trunk}.
    \item Apache Maven, from \url{http://maven.apache.org/index.html}
\end{itemize}

The DUCC webserver server optionally supports direct ``jconsole'' attach to DUCC job processes.  To install
this, the following is required:
\begin{itemize}
    \item Apache Ant, any reasonably current version.
\end{itemize}
    
To (optionally) build the documentation, the following is also required:
\begin{itemize}
  \item Latex, including the \emph{pdflatex} and \emph{htlatex} packages.  A good place
    to start if you need to install it is \url{https://www.tug.org/texlive/}.
\end{itemize}

More detailed one-time setup instructions for source-level builds via subversion can be found here:
\url{http://uima.apache.org/one-time-setup.html#svn-setup}

\section{Building from Source}

To build from source, ensure you have
Subversion and Maven installed.  Extract the source from the SVN repository named above. 

Then from your extract directory into
the root directory (usually current-directory/trunk), and run the command
\begin{verbatim}
    mvn install
\end{verbatim}
or
\begin{verbatim}
    mvn install -Pbuild-duccdocs
\end{verbatim}
if you have LaTeX insalled and wish to do the optional build of documentation.
The build-duccdocs profile can also activated if the environment valiable BUILD\_DUCCDOCS is set true.

If this is your first Maven build it may take quite a while as Maven downloads all the
open-source pre-requisites.  (The pre-requisites are stored in the Maven repository, usually
your \$HOME/.m2).

When build is complete, a tarball is placed in your current-directory/trunk/target
directory.

\section{Documentation}
\begin{sloppypar}
After installation the DUCC documentation is found (in both PDF and HTML format) in the directory 
ducc\_runtime/docs.  As well, the DUCC webserver contains a link to the full documentation on each major page.
The API is documented only via JavaDoc, distributed in the webserver's root directory 
{\tt \duccruntime/webserver/root/doc/api.}  
\end{sloppypar}

If building from source, Maven places the documentation in
\begin{itemize}
    \item {\tt trunk/uima-ducc-duccdocs/target/site} (main documentation), and 
    \item {\tt trunk/target/site/apidocs} (API Javadoc)
\end{itemize}

\section{Single System Installation and Verification}

Although any user ID can be used to run a single-system DUCC, creating a ``ducc'' userid is recommended
to enable the later use of cgroups as well as running processes with the credentials of the submitting user.

If multiple nodes are going to be added later, the ducc runtime tree should be installed
on a shared filesystem so that it can be mounted on the additional nodes.

Verification submits a very simple UIMA pipeline for execution under DUCC.  Once this is shown to be
working, one may proceed installing additional features.


\section{Minimal Hardware Requirements for Single System Installation}
\begin{itemize}
    \item One Intel-based or IBM Power-based system (Big or Little Endian).  (More systems may be added later.)

    \item 8GB of memory.  16GB or more is preferable for developing and testing applications beyond
      the non-trivial.  

    \item 1GB disk space to hold the DUCC runtime, system logs, and job logs.  More is
      usually needed for larger installations.  
\end{itemize}

Please note: DUCC is intended for scaling out memory-intensive UIMA applications over computing
clusters consisting of multiple nodes with large (16GB-256GB or more) memory.  The minimal
requirements are for initial test and evaluation purposes, but will not be sufficient to run actual
workloads.

\section{Single System Installation}
\label{subsec:install.single-user}
    \begin{enumerate}
      \item Expand the distribution file with the appropriate umask:
\begin{verbatim}
(umask 022 && tar -zxf <distribution.file>)
\end{verbatim}

        This creates a directory with a name of the form ``apache-uima-ducc-[version]''.
  
        This directory contains the full DUCC runtime which
        you may use ``in place'' but it is highly recommended that you move it
        into a standard location on a shared filesystem; for example, under ducc's HOME directory:
\begin{verbatim}
mv apache-uima-ducc-[version] /home/ducc/ducc_runtime
\end{verbatim}

        We refer to this directory, regardless of its location, as \duccruntime. For simplicity,
        some of the examples in this document assume it has been moved to /home/ducc/ducc\_runtime.

      \item Change directories into the admin sub-directory of \duccruntime: 
\begin{verbatim}
cd $DUCC_HOME/admin
\end{verbatim}

        \item Run the post-installation script: 
\begin{verbatim}
./ducc_post_install
\end{verbatim}
          If this script fails, correct any problems it identifies and run it again.

          Note that {\em ducc\_post\_install} initializes various default parameters which 
          may be changed later by the system administrator.  Therefore it usually should be
          run only during this first installation step.

        \item If you wish to install jconsole support from the webserver, make sure Apache Ant
          is installed, and run
\begin{verbatim}
./sign_jconsole_jar
\end{verbatim}
          This step may be run at any time if you wish to defer it.

   \end{enumerate}

That's it, DUCC is installed and ready to run. (If errors were displayed during ducc\_post\_install
they must be corrected before continuing.)

\section{Initial System Verification}

Here we verify the system configuration, start DUCC, run a test Job, and then shutdown DUCC.

To run the verification, issue these commands.
\begin{enumerate}
  \item cd \duccruntime/admin 
  \item ./check\_ducc
  
    Examine the output of check\_ducc.  If any errors are shown, correct the errors and rerun
    check\_ducc until there are no errors.  
  \item Finally, start ducc: ./start\_ducc
  \end{enumerate}
  
  Start\_ducc will first perform a number of consistency checks.
  It then starts the ActiveMQ broker, the DUCC control processes, and a single DUCC agent on the
  local node.

  You will see some startup messages similar to the following:

\begin{verbatim}
ENV: Java is configured as: /share/jdk1.7/bin/java
ENV: java full version "1.7.0_40-b43"
ENV: Threading enabled: True
MEM: memory is 15 gB
ENV: system is Linux
allnodes /home/ducc/ducc_runtime/resources/ducc.nodes
Class definition file is ducc.classes
OK: Class and node definitions validated.
OK: Class configuration checked
Starting broker on ducchead.biz.org
Waiting for broker ..... 0
Waiting for broker ..... 1
ActiveMQ broker is found on configured host and port: ducchead.biz.org:61616
Starting 1 agents
********** Starting agents from file /home/ducc/ducc_runtime/resources/ducc.nodes
Starting warm
Waiting for Completion
ducchead.biz.org Starting rm
     PID 14198
ducchead.biz.org Starting pm
     PID 14223
ducchead.biz.org Starting sm
     PID 14248
ducchead.biz.org Starting or
     PID 14275
ducchead.biz.org Starting ws
     PID 14300
ducchead.biz.org
    ducc_ling OK
    DUCC Agent started PID 14325
All threads returned
\end{verbatim}

  Now open a browser and go to the DUCC webserver's url, http://$<$hostname$>$:42133 where $<$hostname$>$ is
  the name of the host where DUCC is started.  Navigate to the Reservations page via the links in
  the upper-left corner.  You should see the DUCC JobDriver reservation in state
  WaitingForResources.  In a few minutes this should change to Assigned.
  Now jobs can be submitted.
  
  To submit a job,
  \begin{enumerate}
    \item \duccruntime/bin/ducc\_submit --specification \duccruntime/examples/simple/1.job
    \end{enumerate}
    
    Open the browser in the DUCC jobs page.  You should see the job progress through a series of
    transitions: Waiting For Driver, Waiting For Services, Waiting For Resources, Initializing, and
    finally, Running.  You'll see the number of work items submitted (15) and the number of work
    items completed grow from 0 to 15.  Finally, the job will move into Completing and then
    Completed..

    Since this example does not specify a log directory DUCC will create a log directory in your HOME directory under 
\begin{verbatim}
$HOME/ducc/logs/job-id
\end{verbatim}

    In this directory, you will find a log for the sample job's JobDriver (JD), JobProcess (JP), and
    a number of other files relating to the job.

    This is a good time to explore the DUCC web pages.  Notice that the job id is a link to a set of
    pages with details about the execution of the job.

    Notice also, in the upper-right corner is a link to the full DUCC documentation, the ``DuccBook''.

    Finally, stop DUCC:
    \begin{enumerate}
      \item cd \duccruntime/admin
      \item./stop\_ducc -a
      \end{enumerate}
      

\section{Add additional nodes to the DUCC cluster}
   Additional nodes must meet all 
   \ifdefined\DUCCSTANDALONE
   {\em prerequisites} (listed above).
   \else
   \hyperref[sec:install.prerequisites]{\em prerequisites}.
   \fi

   \$DUCC\_HOME must be on a shared filesystem and mounted at the same location
   on all DUCC nodes.

   If user's home directories are on local filesystems the location for user logfiles
   should be specified to be on a shared filesystem. 

   Addional nodes are normally added to a worker node group. Note that the
   DUCC head node does not have to be a worker node.
   In addition, the webserver node can be separate from the DUCC head node 
   (see webserver configuration options in ducc.properties).

   For worker nodes DUCC needs to know what node group
   each machine belongs to, and what nodes need an Agent process to be started on.

   The configuration shipped with DUCC have all nodes in the same "default" node pool.
   Worker nodes are listed in the file
\begin{verbatim}
$DUCC_HOME/resources/ducc.nodes.  
\end{verbatim}
   
   During initial installation, this file was initialized with the node DUCC is installed on.
   Additional nodes may be added to the file using a text editor to increase the size of the DUCC
   cluster.


\section{Ducc\_ling Configuration - Running with credentials of submitting user}
\label{sec:duccling.install}

   DUCC launches user processes through ducc\_ling, a small native C application.
   By default the resultant process runs with the credentials of the user ID of
   the DUCC application. It is possible for multiple users to submit work to
   DUCC in this configuration, but it requires that the user ID running DUCC has
   write access to all directories to which the user process outputs data.
   By configuring the ducc user ID and ducc\_ling correctly, work submitted by
   all users will run with their own credentials.  

    Before proceeding with this step, please note: 
    \begin{itemize}
        \item The sequence operations consisting of {\em chown} and {\em chmod} MUST be performed
          in the exact order given below.  If the {\em chmod} operation is performed before
          the {\em chown} operation, Linux will regress the permissions granted by {\em chmod} 
          and ducc\_ling will be incorrectly installed.
    \end{itemize}

    ducc\_ling is designed to be a setuid-root program whose function is to run user processes with the identity of
    the submitting user. This must be installed correctly; incorrect installation can prevent jobs from running as
    their submitters, and in the worse case, can introduce security problems into the system.

    ducc\_ling can either be installed on a local disk on every system in the DUCC cluster, 
    or on a shared-filesystem that does not suppress setuid-root permissions on client nodes.
    The path to ducc\_ling must be the same on each DUCC node. 
    The default path configuration is
    \${DUCC\_HOME}/admin/\$\{os.arch\}/ in order to handle clusters with mixed OS platforms.
    \$\{os.arch\} is the architecture specific value of the Java system property with that name;
    examples are amd64 and ppc64.
   
	The steps are: build ducc\_ling for each node architecture to be added to the cluster,
	copy ducc\_ling to the desired location, and then configure ducc\_ling to give user
	ducc the ability to spawn a process as a different user.

    In the example below ducc\_ling is left under \$DUCC\_HOME, where it is built.

    As user {\em ducc}, build ducc\_ling for necessary architectures (this is done
    automatically for the DUCC head machine by the ducc\_post\_install script).
    For each unique OS platform:
    \begin{enumerate}
        \item cd \$DUCC\_HOME/admin
        \item ./build\_duccling
     \end{enumerate}

     Then, as user {\em root} on the shared filesystem, {\em cd \$DUCC\_HOME/admin}, and for each unique OS architecture:
     \begin{enumerate}
        \item chown ducc.ducc \$\{os.arch\}
        \\ (set directory ownership to be user ducc, group ducc)
        \item chmod 700 \$\{os.arch\}
        \\ (only user ducc can read contents of directory)
        \item chown root.ducc \$\{os.arch\}/ducc\_ling
        \\ (make root owner of ducc\_ling, and let users in group ducc access it)
        \item chmod 4750 \$\{os.arch\}/ducc\_ling
        \\ (ducc\_ling runs as user root when started by users in group ducc)
     \end{enumerate}
          
If these steps are correctly performed, ONLY user {\em ducc} may use the ducc\_ling program in
a privileged way. ducc\_ling contains checks to prevent even user {\em root} from using it for
privileged operations.

If a different location is chosen for ducc\_ling the new path needs to be specified 
for ducc.agent.launcher.ducc\_spawn\_path in \$DUCC\_HOME/resources/site.ducc.properties.
For more information see
\ifdefined\DUCCSTANDALONE
{\em ``Properties merging''} in the DuccBook. 
\else
\hyperref[sec:admin.properties-merge] {Properties merging}. 
\fi


\section{CGroups Installation and Configuration}

\begin{description}
    \item[Note:] A key feature of DUCC is to run user processes in CGroups in order to guarantee
      each process always has the amount of RAM requested. RAM allocated to the managed process
      (and any child processes) that exceed requested DUCC memory size will be forced into swap space. 
      Without CGroups a process that exceeds its requested memory size by N\% is killed 
      (default N=5 in ducc.properties), and memory use by child processes is ignored.
      
      DUCC's CGroup configuration also allocates CPU resources to managed processes based on
      relative memory size. A process with 50\% of a machine's RAM will be guaranteed at least
      50\% of the machine's CPU resources as well. 
\end{description}

    The steps in this task must be done as user root and user ducc.

    To install and configure CGroups for DUCC:
    \begin{enumerate}
       \item Install the appropriate 
   \ifdefined\DUCCSTANDALONE
   libcgroup package at level 0.37 or above (see {\em Installation Prerequisites}).
   \else
   \hyperref[sec:install.prerequisites]{libcgroup package} at level 0.37 or above.
   \fi

       \item Configure /etc/cgconfig.conf as follows:
\begin{verbatim}
   # Mount cgroups for older OS (e.g. RHEL v6)
   # For newer OS, remove entire mount block 
   mount {
      cpuset = /cgroup/cpuset;
      cpu = /cgroup/cpu;
      cpuacct = /cgroup/cpuacct;
      memory = /cgroup/memory;
      devices = /cgroup/devices;
      freezer = /cgroup/freezer;
      net_cls = /cgroup/net_cls;
      blkio = /cgroup/blkio;
   }
   # Define cgroup ducc and setup permissions
   group ducc {
    perm {
        task {
           uid = ducc;
        }
        admin {
           uid = ducc;
        }
    }
    memory {}
    cpu{}
   }
\end{verbatim}
       \item Start the cgconfig service:
\begin{verbatim}
   service cgconfig start
\end{verbatim}
         
       \item Verify cgconfig service is running by the existence of directory: 
\begin{verbatim}
   /cgroups/ducc
\end{verbatim}

       \item Configure the cgconfig service to start on reboot:
\begin{verbatim}
   chkconfig cgconfig on
\end{verbatim}
    \end{enumerate}

{\em Note:} if CGroups is not installed on a machine the DUCC Agent will detect this and not 
  	attempt to use the feature. CGroups can also be disabled for all machines
   	(see 
   \ifdefined\DUCCSTANDALONE
   {\em ducc.agent.launcher.cgroups.enable} in ducc.properties, described in the DuccBook.)
   \else
   \hyperref[itm:props-agent.cgroups.enable] {\em ducc.agent.launcher.cgroups.enable}) 
   \fi
   	or it can be disabled for individual machines (see 
   \ifdefined\DUCCSTANDALONE
   {\em ducc.agent.exclusion.file} in ducc.properties, described in the DuccBook.)
   \else
   \hyperref[itm:props-agent.cgroups.exclusion]{\em ducc.agent.exclusion.file}).
   \fi


 
\section{Full DUCC Verification}

This is identical to initial verification, with the one difference that the job ``1.job'' should be
submitted as any user other than ducc.  Watch the webserver and check that the job executes
under the correct identity.  Once this completes, DUCC is installed and verified.
 
\section{Enable DUCC webserver login}

    This step is optional.  As shipped, the webserver is disabled for
    logins.  This can be seen by hovering over the Login text located in the
    upper right of most webserver pages: 
\begin{verbatim}
System is configured to disallow logins
\end{verbatim}

    To enable logins, a Java-based authenticator must be plugged-in and the
    login feature must be enabled in the ducc.properties file by the DUCC
    administrator.  Also, ducc\_ling should be properly deployed (see 
    Ducc\_ling Installation section above).
    
    A beta version of a Linux-based authentication plug-in is shipped with DUCC.
    It can be found in the source tree:
\begin{verbatim}
org.apache.uima.ducc.ws.authentication.LinuxAuthenticationManager
\end{verbatim}

    The Linux-based authentication plug-in will attempt to validate webserver
    login requests by appealing to the host OS.  The user who wishes to
    login provides a userid and password to the webserver via https, which
    in-turn are handed-off to the OS for a success/failure reply.
    
    To have the webserver employ the beta Linux-based authentication plug-in,
    the DUCC administrator should perform the following as user ducc:
\begin{verbatim}    
1. edit ducc.properties
2. locate: ducc.ws.login.enabled = false
3. modify: ducc.ws.login.enabled = true
4. save
\end{verbatim}

    Note: The beta Linux-based authentication plug-in has limited testing.
    In particular, it was tested using:
\begin{verbatim}
Red Hat Enterprise Linux Workstation release 6.4 (Santiago)
\end{verbatim}    
    
    Alternatively, you can provide your own authentication plug-in.  To do so:
\begin{verbatim}    
1. author a Java class that implements 
   org.apache.uima.ducc.common.authentication.IAuthenticationManager
2. create a jar file comprising your authentication class
3. put the jar file in a location accessible by the DUCC webserver, such as 
    $DUCC_HOME/lib/authentication
4. put any authentication dependency jar files there as well
5. edit ducc.properties
6. add the following:
   ducc.local.jars = authentication/*
   ducc.authentication.implementer=<your.authenticator.class.Name>
7. locate: ducc.ws.login.enabled = false
8. modify: ducc.ws.login.enabled = true
9. save   
\end{verbatim}