SERVER INSTALLATION GUIDE PIPELINE PILOT SERVER 2016 Copyright Notice

©2015 Dassault Systèmes. All rights reserved. 3DEXPERIENCE, the Compass icon and the 3DS logo, CATIA, SOLIDWORKS, ENOVIA, DELMIA, SIMULIA, GEOVIA, EXALEAD, 3D VIA, BIOVIA and NETVIBES are commercial trademarks or registered trademarks of Dassault Systèmes or its subsidiaries in the U.S. and/or other countries. All other trademarks are owned by their respective owners. Use of any Dassault Systèmes or its subsidiaries trademarks is subject to their express written approval.

Acknowledgments and References

To print photographs or files of computational results (figures and/or data) obtained using BIOVIA software, acknowledge the source in an appropriate format. For example: "Computational results obtained using software programs from Dassault Systèmes BIOVIA. The ab initio calculations were performed with the DMol3 program, and graphical displays generated with BIOVIA Pipeline Pilot Server."

BIOVIA may grant permission to republish or reprint its copyrighted materials. Requests should be submitted to BIOVIA Support, either through electronic mail to [email protected], or in writing to:

BIOVIA Support 5005 Wateridge Vista Drive, San Diego, CA 92121 USA Contents

Chapter 1: Pipeline Pilot Server 2016 Post-Install Checklist 27 Installation 5 Pipeline Pilot Server Home Page 30 Quick Install Tips 5 Administration Portal 31 BIOVIA License Files 5 Chapter 7: Deployment Options 32 Additional Information 5 Load Balancing Deployments 32 Chapter 2: Getting Started 6 Load Balancing Features 32 Overview 6 Pre-configuration Steps 32 Deployment Options 6 Configuration Steps 32 Managing Multiple Installations 6 Grid Engine Installation 34 Side-by-side installations 6 Grid Engine setup 34 Protocol database compatibility 7 Grid Engine requirements 34 Port management 7 Setting the User Account for Running Xvfb Sharing an XMLDB with Multiple Servers 7 on Grid Cluster Nodes 34 Moving/Migrating an Installation 8 Chapter 8: Uninstalling Pipeline Pilot 36 Best practices for moving directories 8 Uninstalling from a Windows Server 36 Pre-Install Checklist 10 Uninstall from a Server 37 Chapter 3: Pipeline Pilot Services 12 Appendix A: Unattended Installation on Shutting down Services on Windows 13 Linux 39 Shutting down Apache on Linux 14 Setting Required Pre-Installation Environment Variables 39 Chapter 4: Installing Pipeline Pilot on Windows 15 Installation Environment Variables 40 Windows Server Installation 15 Package names. 40 Windows Guidelines 16 Installation Command Example 42 Windows Requirements for Load Balancing 16 Appendix B: Configuring Apache as a Load Balancer and Reverse Proxy 44 Local job temporary directory 16 HTTP Requests 44 Impersonation on Windows 17 SSL Requests 45 Chapter 5: Installing Pipeline Pilot on Linux 18 Appendix C: Installing Separately Licensed Pre-Installation Checklist for Linux 18 Collections 46 Linux Guidelines 20 Materials Studio Collection 46 Installation paths 20 Windows (upgrade) 46 Red Hat Linux 20 Windows (new installation) 46 Local authentication and SELinux 20 Linux (upgrade) 47 Distributed file system and NFS Linux (new installation) 47 configuration 20 Appendix D: Required Linux RPMs 48 NFS Settings 21 Red Hat Enterprise Linux 7 48 NFS Cache Settings 21 Red Hat Enterprise Linux 6 49 Clock synchronization 21 SUSE Linux Enterprise 11 50 Possible interaction with /usr/sbin/tmpwatch 22 Installing Pipeline Pilot on Linux 22 Post-Install Checklist for Linux 23 Linux Checklist 23 Server Configuration Checklist 24 Testing Client software on Linux 26 Chapter 6: Post-Installation 27 Chapter 1: Pipeline Pilot Server 2016 Installation

Quick Install Tips

To install Pipeline Pilot on your Windows or Linux server 1. Shut Down Apache. 2. Review Pre-Install Checklist and perform all recommended pre-installation tasks. 3. Follow the instructions for Windows Server Installation or Linux Server Installation. 4. You might need to perform additional tasks on your server to ensure your installation runs properly. For details, see Post-Install Checklist.

BIOVIA License Files To install any BIOVIA Pipeline Pilot Server software and collections, you need a valid license file. Instructions are included with the license file activating a new license. Note that the Thomson Reuters and Materials Studio collections require a separate license. See Appendix C Installing Separately Licensed Collections.

Additional Information For information about supported hardware and software, see the Pipeline Pilot System Requirements document. For details about advanced server deployments, see the Pipeline Pilot Deployment Guide. To install Pipeline Pilot Client, see the Pipeline Pilot Client Guide. For details about installing Software, see the R Software Installation and Configuration Guide. For more information about BIOVIA Pipeline Pilot Server 2016 and other Pipeline Pilot software products, visit https://community.3dsbiovia.com/.

Pipeline Pilot Server 2016 Installation | Page 5 Chapter 2: Getting Started

Overview Upfront planning is critical to a successful enterprise deployment. See the Pipeline Pilot Deployment Guide, then use the pre-installation checklists and related information in this chapter to plan your deployment. The checklists apply to most deployments. If you have any questions, contact [email protected]. IMPORTANT! Pipeline Pilot Server 2016 requires Pipeline Pilot Client 2016. Users must upgrade their client in order to access the Server. Some applications and collections must be upgraded separately (e.g., Insight and Insight for Excel). Before installing Pipeline Pilot, ensure that you have access to these separate installers that are compatible with this version of the server. BIOVIA Draw does not upgrade automatically when you upgrade from an earlier version of Pipeline Pilot Server. You will need to upgrade BIOVIA Draw automatically. A separate license file is required for the Thomson Reuters and Materials Studio collections. See Appendix C: Installing Separately Licensed Collections.

Deployment Options There are several advanced options for deploying Pipeline Pilot including: Workstation Standalone Web Farm (Load balancing) Clustering (Linux only) Distributed grid computing (Linux only) Note: For more information, see Chapter 7: Deployment Options

Managing Multiple Installations Pipeline Pilot supports side-by-side installations of existing and new products. You can run BIOVIA Pipeline Pilot Server 2016 and previous Pipeline Pilot Server 9.x versions on the same machine. With side-by-side installations, the installer can copy components and protocols from an existing protocol database (XMLDB) into the new installation. You can also upgrade an existing installation to the latest version. Side-by-side installations To run different versions side-by-side, each version needs to be installed in a unique path or drive on your server. They cannot all reside in the same directory or subdirectories. You also need to run side-by- side versions of the clients that correspond to each server installation, assuming they are mutually incompatible versions.

Getting Started | Page 6 On a Windows server, the default paths are:

Version Default install path Pipeline Pilot Server 2016 C:\Program Files\BIOVIA\PPS Pipeline Pilot Server 9.5 C:\Program Files\BIOVIA\PPS Pipeline Pilot Server 9.0, 9.1, and 9.2 C:\Program Files\Accelrys\AEP

Pipeline Pilot Server 8.x C:\Program Files\Accelrys\PPS

Pipeline Pilot Server 7.5 C:\Program Files\Accelrys\SES

Pipeline Pilot 7.0 and earlier versions: C:\Program Files\SciTegic (or "SciTegic 7.0" if this location already contains an installation)

Protocol database compatibility The Pipeline Pilot Client (Pro Client) supports connecting to multiple XMLDBs through the Network tab. The 2016, 9.x and 8.x clients can connect to databases of previous software versions, but can only save files on the current database. Port management The following port numbers used by Pipeline PIlot: Type Default Port HTTP 9944 HTTPS 9943 Tomcat Shutdown 9945 Tomcat HTTP 9946 Derby 9947 You can have side-by-side installations of Pipeline Pilot 2016 on the same server where Pipeline Pilot 9.5 and earlier versions are installed. Each installation requires unique ports. If the default ports listed above are already in use for 2016, 9.1, 9.0, 8.5, or an earlier version of Pipeline Pilot Server, use alternate port numbers.

Sharing an XMLDB with Multiple Servers Pipeline Pilot Pro clients running at multiple sites can access the same protocols and components if they share the same protocol database (XMLDB). To provide this functionality, the computational servers that process protocol jobs at each site need to communicate with one server that operates as the XMLDB for all the other servers. Clients are only aware of the computational server where they run protocols. They have no information about the machine hosting the XMLDB. See Moving/Migrating an Installation for more information about XMLDB.

Page 7 | Pipeline Pilot Server • Server Installation Guide Moving/Migrating an Installation On Windows, you can move/migrate your server location by uninstalling Pipeline Pilot and transferring the XMLDB to the new server. After uninstalling, some files remain intact, making it possible for you to relocate existing data to a server re-installation. Specifically, your XMLDB remains intact in the folder, so you can easily move/migrate it to a new location. Best practices for moving directories

Note: is the name of your Pipeline Pilot Server installation path, for example, C:\Program Files\BIOVIA\PPS.

Default Directory Description When moving your server to a new location, do this: /xmldb Protocol database that Copy this directory to the target stores protocols and location, so users can access their components specific to the protocols, components, and version server, including version history. history and custom protocols. Some files are public (Components and Protocols subfolders); others are owned by specific users (User subfolder). /xmldb/objects This information should only the objects folder to the target be located where Pipeline server. Instead, use the Admin Pilot is installed. Contents Portal to customize your server are configured on your settings on the target server. This server during installation ensures that all registry files are and must always reside in correctly configured. the default directory. /web Custom web applications, When moving your server to a new Web server customizations, location, you need these files on the job folders. target server, so users can continue to run your web applications. /apps/ This directory is removed, NA scitegic but any other non-BIOVIA packages in /apps are not deleted. /licensing License-related files. When moving your server to a new location, copy this folder to the target location. During installation, specify the path to your license file to install the appropriate packages.

Getting Started | Page 8 /public If this directory contains NA custom user data, including /public/users, it is not removed. /logs Log files. You do not need to copy these files to other servers as new log files are created after installing. Move this folder to preserve the server history, (for usage tracking purposes). /install Contains installation files. You do not need to copy these files to other servers.

Page 9 | Pipeline Pilot Server • Server Installation Guide Pre-Install Checklist To ensure that your Pipeline Pilot installation goes smoothly, complete the checklist below. If you have any questions, contact [email protected]. IMPORTANT! If you are installing on a machine with an older version, it is recommended that you backup your components and protocols before installing. See step 13 in the checklist.

Do this Details Done 1. Obtain required system Windows: System Administrator administrator privileges to Linux: Do not install as Root install software on your server. 2. Locate your BIOVIA license file A valid license file is current (not expired) and (*.lic). accessible during the installation. For each new license file that you obtain, instructions are available that explain how to update and activate the license file. BIOVIA sends the license file to the Ship To contact for your company. Note: If you are planning to use the Thomson Reuters or Materials Studio collection, you will need to perform additional installation steps after the initial installation. See Appendix C Installing Separately Licensed Collections.

3. Verify that your server meets See the Pipeline Pilot System Requirements the hardware/software document. requirements. 4. Check to see if any protocols are Active jobs on the server must finish running before currently running on your installing Pipeline Pilot. Cancel jobs in the Admin server. Portal (Jobs > Running Jobs). 5. Shut down Apache. Apache cannot be running on the server where you are installing Pipeline Pilot. See Pipeline Pilot Services. 6. Shut down the Java server. If your server also runs the Java Server package (optional), Apache Tomcat cannot be running when you install Pipeline Pilot. See Pipeline Pilot Services. 7. Close all instances of the Admin Shut down all Pipeline Pilot and all Pro Client Portal and Pipeline Pilot Client, applications on the server. Close all Pipeline Pilot and Web Port clients. clients with current connections to the protocol database (XMLDB) on the server you are about to upgrade. 8. Disable all virus scanning Some virus scanners and firewall applications can software and firewalls on the slow down or even corrupt the installation. server.

Getting Started | Page 10 9. Verify that you can use the If these port numbers are already in use by another Pipeline Pilot default port web service, you can manually assign different port numbers with the Apache web numbers for Apache during the installation. server: Decide in advance what port numbers to use. You 9944: HTTP need to refer to these port numbers when accessing 9943: HTTPS the Home Page and server-related applications. Make 9945: Tomcat Shutdown a note of any custom port numbers you decide to manually assign. 9946: Tomcat HTTP The HTTP and HTTPS port numbers are defined during 9947: Derby Port installation. The other three are automatically selected during installation. However, you can explicitly define all port settings. All use of Tomcat specific ports is within the server node, so they should be externally blocked to increase protection. See the Admin Portal Help Center: Admin > Setup and Configuration > Server Configurations > Reconfiguring Ports. 10. Do you need to run the Apache If yes, you need to obtain the specific user account Web Server under a specific user and password. to extend and control local network resources? 11. Are you installing on a Security- Run "sciseallow" as root before installing Pipeline Enhanced Linux (SELinux) Pilot. machine? 12. Are you installing on a grid, If yes, configure and test your servers to ensure they cluster, web farm, or reverse are fully operational before installing Pipeline Pilot. proxy? For details, see the Pipeline Pilot Deployment Guide. 13. Are you installing on a machine If yes: with an older version or Windows reinstalling and need to make a 1. Open a command prompt. backup of your components 2. Navigate to /bin. and protocols (xmldb)? 3. Run the following command: DbUtil.exe - backup -file xmldb_backup.zip. This will create the following file: /public/backups/xmldb_ backup.zip. Linux 1. Open a command prompt. 2. Run the following command: /linux_bin/ppvars.sh/linux_bin/DbUtil -backup -file xmldb_backup.zip.

Page 11 | Pipeline Pilot Server • Server Installation Guide Chapter 3: Pipeline Pilot Services

All Pipeline Pilot services (including Apache HTTPD and Java/Tomcat) need to be shut down before you can install a new version of Pipeline Pilot on your server. This frees up the ports, which is necessary for a successful installation. The default port numbers used by the Apache web service are: Protocol Default Port HTTPS 9943 HTTP 9944 Tomcat Shutdown 9945 Tomcat HTTP 9946 Derby 9947 The following services are installed with your server installation on Windows:

BIOVIA Pipeline Pilot Server BIOVIA Pipeline Pilot 16.1.0 (Manager): Manages the lifetime of the 2016 Services Pipeline Pilot Server environment. BIOVIA Pipeline Pilot 16.1.0 Service (Httpd): Service that runs the Apache web server. BIOVIA Pipeline Pilot 16.1.0 Service (Tomcat): Service that runs the Tomcat Java application server. BIOVIA Pipeline Pilot 9.5.0 BIOVIA Pipeline Pilot 9.5.0 (Manager): Manages the lifetime of the Services Pipeline Pilot Server environment. BIOVIA Pipeline Pilot 9.5.0 Service (Httpd): Service that runs the Apache web server. BIOVIA Pipeline Pilot 9.5.0 Service (Tomcat): Service that runs the Tomcat Java application server. Pipeline Pilot Server 9.0, 9.1, Accelrys Enterprise Platform Service 9_x: Standard AEP service that 9.2 Services runs on the server. AEP 9.x (Httpd): Service that runs your Apache web server. AEP 9.x (Tomcat): Service that runs the Tomcat Java application server. Pipeline Pilot Server 8.5 and scitegic_apache_ x_x_x earlier scitegic_derby_ x_x_x scitegic_tomcat_ x_x_x

Pipeline Pilot Server Services | Page 12 Notes: The installer can shut down these services for you while you perform a new Pipeline Pilot installation. You can also use the Windows Services console to shut down services the same way you shut down web services with earlier Pipeline Pilot Server releases. For Pipeline Pilot Server 8.5 and earlier, you need to manually shut down the Apache web server before you can install. If multiple Pipeline Pilot servers are running simultaneously, multiple Apache services are also running on your server and it is necessary to shut down all currently running services before performing an installation.

Shutting down Services on Windows 1. From Windows Control Panel, go to Administrative Tools > Services. The Services console opens. 2. To shut down all services, right click the BIOVIA Pipeline Pilot 9.5 (Manager) service in 9.5 (or the Accelrys Enterprise Platform Service 9_x for 9.0, 9.1, or 9.2) and select Stop. As the manager service, stopping this service will also stop the Httpd and Tomcat services. To shut down either the Httpd or Tomcat services individually, right-click on the appropriate entry and select Stop. 3. If you want to ensure that all ports are released, open a command window and run the following: netstat -a A list of port numbers currently in use is displayed. This list should not include Apache ports. Tip: It can take a few minutes for Apache to release the ports. Retype the command to refresh the list.

Page 13 | Pipeline Pilot Server • Server Installation Guide Shutting down Apache on Linux To shut down Apache on a Linux server: 1. Change directories to your directory: $ cd linux_bin 2. Stop the server: $./stopserver 3. Ensure that the ports are released: $ netstat -lt A list of port numbers currently in use is displayed. The default Apache port numbers used by Pipeline Pilot should not be listed.

Tips: It might take a few minutes for Apache to release the ports. To refresh the list, retype the command. You also can use the following command to check for running HTTPD processes in the scitegic directory: $ ps -aux | grep scihttpd

Pipeline Pilot Server Services | Page 14 Chapter 4: Installing Pipeline Pilot on Windows

Windows Server Installation To download the Pipeline Pilot files 1. Before installing, review the Pre-Install Checklist. 2. To download individual files, click the Download link. To use the Download Manager, select the check boxes and choose Download Selected Files. 3. Save the .7z file to your computer and unzip it. To install Pipeline Pilot on a Windows server 1. From the folder that contains your installation files, run: BPPF2016\bin\scitegicsetup.exe. 2. When the Pipeline Pilot installer starts, review the Welcome screen and then click Next. 3. If prompted, install the required Microsoft runtime libraries (DLLs). 4. At Select a Destination Directory, accept the default C:\Program Files\BIOVIA\PPS or click Change and navigate to the location where you want Pipeline Pilot installed, and then click Next. 5. Browse to or enter the path to your license file (*.lic). It determines what software and component collections to install on your server. Note: The Pipeline Pilot license file was sent via email to the Ship To contact for your company. For easy access, this license file should be copied to the server where you plan to install Pipeline Pilot. 6. At the Select products to install dialog, the products lists are collapsed. Expand the list to view which products are included with your Pipeline Pilot license. Click Next to install all the products. You can uncheck any product that you do not want installed. Review the list carefully. By default, the installer backs up your XMLDB in root/public/backups (in compressed zip format).

Notes: The Platform product is installed by default. To install a collection update, click Add/Remove Packages after the installer starts. To reinstall the same version again, click Reinstall/Repair. 7. If the WARNING! Running Applications window appears, click Stop Services to close the services that appear in the dialog window. 8. Verify the port numbers for the Apache web server. The following are the default port numbers: Type Default Port HTTP 9944 HTTPS 9943 Tomcat Shutdown 9945 Tomcat HTTP 9946

Installing Pipeline Pilot Server on Windows | Page 15 Type Default Port Derby 9947 If another service is already using these ports, you can assign alternate ports for HTTP and HTTPS during installation. Be sure to make a note of what you change so can easily refer to these ports later. To reconfigure the other ports after installation, see the Admin Portal Help Center: Admin > Setup and Configuration > Server Configurations > Reconfiguring Ports. 9. At Confirm Your Installation, click Next to start the installation. 10. If you want to view each package as it installs, check Show installation progress in command window. 11. At Confirm Apache Login, click Next if you do not want to modify the Apache user. To modify the Apache user, click Apache and edit the Log On property for your Apache service. 12. At Installation Complete, click Finish to exit the installer. 13. If you want to open the Pipeline Pilot Server Home Page, select the check box before you click Finish.

Notes: If there is a problem during the installation, the last screen of the installer provides an option for viewing your installation log. It provides details about what took place during the installation, to help you pinpoint the nature of the problem. The file is saved in: \logs\install. After Pipeline Pilot is installed, several services are started. See Pipeline Pilot Server Services. You can stop and restart these services from the Windows Services console. See Shutting Down Apache on Windows After Pipeline Pilot is installed, you have the option to install Pipeline Pilot Client and/or Data Source Builder on the server. The Pipeline Pilot Home Page also has an option for installing Pipeline Pilot Client and/or Data Source Builder (for details, see Post-Install Checklist). The default Home Page URL is: http://: (for example, http://myserver:9944). Client users can install Pipeline Pilot Client by connecting to the Server Home Page from their Web browsers. Users can also read the Pipeline Pilot Client Installation Guide to get further details on Windows client installation.

Windows Guidelines Windows Requirements for Load Balancing See Load Balanced Server (Web Farm) Installation, or the Pipeline Pilot Deployment Guide. Local job temporary directory For best performance, it is possible to use a local disk for temporary files generated while a job is running. This minimizes network traffic and reduces job run time. The local job temp directory must exist on all nodes where the server is running in the exact same path. Because some protocols generate a great deal of temporary data, this directory should be located on a file system that is large enough to hold the data. For more information Jobs, see the Pipeline Pilot Deployment Guide.

Page 16 | Pipeline Pilot Server • Server Installation Guide Impersonation on Windows For full impersonation, configure the Apache service (BIOVIA Pipeline Pilot 16.1.0 Service (Httpd)) to run under the local system account. Alternately, you can configure the server to run under an account that has administrative privileges on the server. The account must also be configured with the Windows Local Security Policy rights Replace a process level token and Adjust memory quotas for a process. If you are configuring for a load balancing deployment, use an account with administrative privileges. For more information on impersonation, see the Pipeline Pilot Deployment Guide.

Installing Pipeline Pilot Server on Windows | Page 17 Chapter 5: Installing Pipeline Pilot on Linux

Pre-Installation Checklist for Linux Review the checklist below to ensure that you perform all preliminary tasks before deploying the Pipeline Pilot Pipeline Pilot Serverin a Linux distributed environment.

Do this Details Done 1. If you are installing on a See Local authentication and SELinux. Security-Enhanced Linux (SELinux) machine, run "sciseallow" as root before installing Pipeline Pilot. 2. Configure a user and group This account should also be available on the cluster. The account for running Pipeline Pipeline Pilot user account does not require a login. Normal Pilot. users should not be members of this group. This account group should only be used for this user. No other users should belong to this group. This can be a no-login account. When deploying on a cluster or grid, the account must be accessible from all machines. 3. To increase performance Making this change requires restarting NFS: and minimize I/O 1. Increase RPCNFSDCOUNT from 8 to a minimum of 16. bottlenecks, increase the On Red Hat servers this value is changed by editing the number of NFS daemons line defining RPCNFSDCOUNT in the file running on the primary /etc/rc.d/init.d/nfs while logged in as root. Pipeline Pilot. 2. Restart the NFS service to enact the change: /etc/init.d/nfs restart 4. Ensure that you have The primary Pipeline Pilot Server should be installed on a sufficient local disk space to disk that is local to the machine that runs it. Installing on install Pipeline Pilot an NFS mounted disk might result in protocol failures and software and related data. other issues. See Distributed file system and NFS This also applies to the local configuration. job temporary directory. To IMPORTANT! The file system mount point path must be use a directory other than the same on all nodes. If you install in /opt/BIOVIA, all /tmp, it is necessary to first cluster nodes need to access the Pipeline Pilot set it up. installation as /opt/BIOVIA. This also applies to load balanced configurations where each node has its own installation.

Installing Pipeline Pilot Server on Linux | Page 18 5. Create the file system For example: mount point. This also # mkdir /opt/BIOVIA applies to load balanced chown ppuser:ppgroup /opt/BIOVIA configurations. 6. Export the file system Edit the file /etc/exports and add entries for each node mount point. you want to export to. The no_root_squash option should be set. /opt/BIOVIA/ Node1(rw,no_root_squash,sync) Node2(rw,no_root_squash,sync) #exportfs –a See Distributed file system and NFS configuration. 7. For each node in the cluster, Edit the file /etc/fstab and add an entry for the mount the exported file exported file system using the following options: system. pp_server:/opt/BIOVIA/opt/BIOVIA nfs rw,hard,intr,lock,noatime,acdirmin=0, acdirmax=0,acregmin=0,acregmax=0 # mount /opt/BIOVIA See Distributed File System and NFC Configuration. 8. If you are installing on a Manually run a job through the system and verify that it grid, verify that you can works. This usually involves using the grid engine qsub submit a simple job to the command. For further information, refer to your grid grid engine and that it runs engine documentation. See also, the Pipeline Pilot successfully. Deployment Guide. 9. If you are using R Statistics For example, to install both R and Discovery Studio and or Discovery Studio®, plan export them with one mount, install Pipeline Pilot in on installing these packages /opt/BIOVIA/Pipeline Pilot, install R in in a location that can be /opt/BIOVIA/R, and install Discovery Studio in exported to the cluster with /opt/BIOVIA/ds. This allows you to export a single share. /opt/BIOVIA so the cluster can access and run all the applications. (See the R Software and Support Guide.) 10. Certain system Before starting the installation, you can display the requirements might be required system libraries by running the check_ required for the correct system.sh script located in the bin/SystemChecker operation of Pipeline Pilot. folder in the installation media. See Appendix D: Required Linux RPMs for a list of required libraries.

Notes: When installing Pipeline Pilot to run in a web farm configuration, install the software in the same location on each machine. (Pipeline Pilot is installed on every node and is not shared.) See Load Balanced Configuration (Web Farm) Installation. For a cluster or grid operation, Pipeline Pilot is installed once. The installation is then shared using the same path on all nodes. For more information on these deployment options, see the Pipeline Pilot Deployment Guide.

Page 19 | Pipeline Pilot Server • Server Installation Guide Linux Guidelines Installation paths Regardless of your environment, when installing Pipeline Pilot in a multiple node configuration, the installation path must be the same on all nodes that access Pipeline Pilot software. For example, if you install the software in /opt/BIOVIA on one machine, the software must be available in /opt/BIOVIA on all machines. This includes installations for load balanced configurations and shared installations such as grids and clusters. For more information on these deployment options, see the Pipeline Pilot Deployment Guide. Red Hat Linux

Tips: If you are installing Pipeline Pilot on Red Hat® Enterprise Linux® v6, install the compat-libstdc++ RPM package. Red Hat no longer ships with Xvfb, which is used by protocols that use R to generate images. RPM for Xvfb is available for download via Red Hat support (see http://www.redhat.com/) .

Local authentication and SELinux If you are using local authentication or planning to use Security-Enhanced Linux (SELinux), it is necessary to perform a few extra steps prior to installing Pipeline Pilot.

Do this Details Done 1. For local password file chgrp ppgroup /etc/shadow authentication, change the chmod g+r /etc/shadow permissions on /etc/shadow so it can Tip: This task is only necessary if you plan on be read. using local accounts.

2. For a machine with SELinux enabled You do not have to disable SELinux to run (Red Hat 5+), run the script that installs Pipeline Pilot. The script is called sciseallow, a module to allow Pipeline Pilot to run. and it is located on the installation disk.

WARNING! Run this script before installing Pipeline Pilot.

Distributed file system and NFS configuration For a clustered or grid environment, the primary server needs to export the installation to the other nodes in the cluster. For load balanced configurations, the installation is not shared, but a shared file system is required for users, jobs, and public directories. Tip: You might get better performance if you use a file system designed for clustering, such as GPFS, Panasas® or Lustre™. The primary requirement for Pipeline Pilot is support for file locking and minimizing directory attribute caching. If you plan on using impersonation with Pipeline Pilot, it is necessary to turn off root squash for NFS mounts.

Installing Pipeline Pilot Server on Linux | Page 20 It is important when dealing with distributed file systems to keep your server clocks synchronized and ensure that the Network Time Protocol daemon (ntpd) is running on all nodes. NFS Settings

Setting Description rw Read and write enabled hard Hard mount option, required by mount. The alternative is soft, which does not work with the Pipeline Pilot. intr System calls are interruptible. Without this option, some operations can cause the server or jobs to lock up while waiting on an NFS system call. lock This is required for the Pipeline Pilot Server. All file operations are controlled by advisory locking. If locking is disabled on the mount, the Pipeline Pilot Server will not work. In some cases where this was omitted for a mount, the server appears to work normally, because lockd is already running on the machine. If there is an existing lockd running, it will handle the lock calls and things will appear to work normally. However, the Pipeline Pilot Server might stop working correctly if the server is rebooted. NFS Cache Settings

Setting Description acdirmin, These settings control attribute caching for directories. Directory attribute acdirmax caching determines the amount of time it takes for a file to show up on other machines. Values specified are in seconds. If directory attribute caching is enabled, operations that depend on a recently written file existing will fail intermittently. The required settings for these are acdirmin=0,acdirmax=0. acregmin, These settings control file attribute caching. File attribute caching determines the acregmax amount of time it takes for file ownership and permission changes to show up on other machines. Values are specified in seconds. The recommended settings for these are acregmin=0,acregmax=0. Note: acregmax should not be set to more than 10 seconds.

rsize, wsize These are the read and write transfer sizes. System defaults should be used. To improve performance, these settings can be safely increased. noatime Disables setting the access time on the file. This is a performance optimization and should be used on all remote mounts.

Tip: A good basic starting point for specifying NFS mount options for the Pipeline Pilot server is: rw,hard,intr,lock,noatime,acdirmin=0,acdirmax=0,acregmin=0,acregmax=0

Clock synchronization In a distributed environment, it is important for all nodes to have synchronized clocks. Ensure that the Network Time Protocol daemon (ntpd) is running on all nodes in the cluster, and that all nodes report the same time.

Page 21 | Pipeline Pilot Server • Server Installation Guide Possible interaction with /usr/sbin/tmpwatch If the server is idle for a period of time, it is possible that the tmpwatch directory cleaner can end up deleting the /var/tmp/scitegic-mutexes-XXXX/.lock file. Under impersonation, this can result in the lock file ending up with the wrong permissions and causing problems on the system. If you are running tmpwatch, edit its configuration file to exclude the /var/tmp/scitegic- mutexes-XXXX directory.

Installing Pipeline Pilot on Linux

IMPORTANT! The installation scripts must include read, write, and execute privileges for the user who is installing. 1. Change the directory to the location of the installation files. This is the directory that contains the sciinstall script. 2. If you are using Red Hat® Enterprise Linux® v5 or later, check for the SELinux extensions: /usr/sbin/selinuxenabled; echo $?

Tip: If "0" displays when the command is executed, and you did not run the sciseallow script as root, do this now. 3. If using impersonation, create an account for this user. The account should use its own group. No regular users should belong to it. 4. Extract the contents from the .tgz file: tar -xpvzf 5. Run sciinstall as a non-root user: $ ./sciinstall 6. Enter 1 to install a new server, and enter a target directory for installing the Pipeline Pilot application. For disk space requirements, see the Pipeline Pilot Server System Requirements document. 7. Edit the path to your BIOVIA license file, including the license name. Your license determines what software and product collections to install on the server. 8. The installer displays the Products that will be installed, the disk space required for the installation, and the available space. Enter y to install Pipeline Pilot and all the available collections. The Platform components are installed by default. If you only want to install Pipeline Pilot and/or specific collections, enter the number of what you do not want installed. Notice that the Action listing for these products has changed to Read Remove or Embed. The embed collection packages will not be removed because they are required for a collection that you want to install. When you have finished your selections, enter y. To install a collection or a collection update, to reinstall an existing installation, or to upgrade an existing installation, rerun the installer and enter a number (1-5) for the action you want to complete. 9. Assign the HTTP and HTTPS ports (required by the Apache web server). The default ports are 9944 (HTTP) and 9943 (HTTPS). Click Y to accept the default port numbers To assign alternate port numbers, press N, and then enter the alternate numbers as the prompt.

Installing Pipeline Pilot Server on Linux | Page 22 Tip: The installer checks the ports to ensure that they are available. If it detects a conflict, a warning is displayed and the installation stops. To assign different port numbers, ensure that they are available before you rerun the installation program. 10. Select the local directory for lock files (that is a directory that is not mounted using NFS). The default directory is /var/tmp. The lock file directory requires, read, write, and execute privileges for all users. For cluster configurations, this directory path must exist on the local hard drive of each node. 11. Enter an option number when asked if the machine you are installing on is the primary node in a cluster. The options are: This node is not in a cluster. Round Robin. Job Leveling

Notes: Depending on your installation choices, it might take some time for the installation to finish. When complete, the Apache and Pipeline Pilot Servers both start automatically. If you did not create a boot script, you need to manually start the server.

Post-Install Checklist for Linux After installing Pipeline Pilot, you can configure features in the Admin Portal to support a particular deployment method. Review the following checklists before you make any changes in the Admin Portal. It might be necessary to perform a few preliminary tasks depending on your operating system and planned deployment. Linux Checklist

Do this Details Done 1. Shut down the Pipeline Pilot For example: Server if it was started during cd /opt/BIOVIA/installdir/linux_bin the installation process. ./stopserver 2. Run scirootinstall on scirootinstall is located in the Linux. apps/scitegic/installer/scripts directory. Even if you do not intend to use impersonation, it's a good idea to run scirootinstall at this time. This script sets up directory permissions, adds init scripts, sets up PAM, and configures the server to run with impersonation. Note: If you are setting up a cluster, scirootinstall should be run on all nodes in the cluster.

Page 23 | Pipeline Pilot Server • Server Installation Guide 3. Set the group bit on Since scirootinstall does this for the default directories on Linux. This installation, this step should only be necessary if you applies when using use an alternate location. alternative locations for the mkdir AltUser users, jobs, or the XMLDB. chown ppuser:ppgroup AltUser This needs to be handled so chmod 2775 AltUser your server can access files created by impersonated users. 4. Run mkpublic. This is mkpublic is located in the required when using a shared apps/scitegic/installer/scripts directory. It public directory in load can be run either as root or the server user. This balanced configurations. command creates a compatible public directory with the correct permissions. Usage: mkpublic [-u user][-g group] path 5. For a grid, run scigridsetup is located in the scigridsetup on each apps/scitegic/installer/scripts directory. node that will run protocols. This script adds an init script for Xvfb and runs sciseallow. IMPORTANT! If you do not run Xvfb, components that use R software to generate plots and graphics might not work. For further details, see the R Software Install and Configuration Guide.

Server Configuration Checklist

Do this Details Done 1. Restart the Linux server if you previously stopped it cd to perform any tasks. /opt/BIOVIA/installdir/ linux_bin ./startserver 2. Log into the Admin Portal and configure the options Features are available for that you want to use. deployment configuration, server security, job management, XMLDB maintenance, etc. 3. If you need to use impersonation, enable it first, and Enable impersonation in the then verify that it works before making other Admin Portal (Security > configuration changes. Authentication). 4. For clustering support, log in to each of the cluster ssh node1 cd nodes and start the servers. Add the nodes to the /opt/BIOVIA/installdir cluster node list and verify that the nodes are active /linux_bin ./startserver (green). Round robin is the recommended clustering method.

Installing Pipeline Pilot Server on Linux | Page 24 5. If you are configuring for a load balancer, export Export and import your server your configuration and save the exported file to configuration in the Admin make it easy to deploy new nodes with the same Portal (Maintenance > Export configuration. Import Configuration). For more information, see Load Balanced Server (Web Farm) Installation.

Page 25 | Pipeline Pilot Server • Server Installation Guide Testing Client software on Linux Pro Client software cannot be installed on a Linux server. To verify the server installation, you must either connect to the server through a client from a separate Windows machine or use a web browser to access the Pipeline Pilot Server Home Page. To access the home page Go to http://:9944 where is the name of your Linux server. Would you like to: Select this: Read documentation about using and administrating Pipeline Pilot Help Center Install Pipeline Pilot Client on Windows Install/Upgrade Client Software Change server settings or enable impersonation Administration Portal Launch the Pipeline Pilot web client Web Port

Installing Pipeline Pilot Server on Linux | Page 26 Chapter 6: Post-Installation

Post-Install Checklist After installing Pipeline Pilot on your server, there might be additional configurations required for connectivity and resource access. BIOVIA recommends that you review the information provided below and identify any outstanding tasks you might need to complete. Note: There are many additional tasks you can perform on the server in addition to the ones listed below. For more detailed information, see the Admin Portal help.

Would you like to: Here is how: Get more familiar with Pipeline Pilot? From your server’s web browser, go to: Be sure to visit the Server Home Page for Pipeline Pilot http://:9944 and launch the Admin Portal, and run example protocols. is the name of your server. Learn about managing Pipeline Pilot? From the Server Home Page, select Help Review the online help available in the Admin Portal, and Center > Administrator Resources. get more familiar with the various ways you can configure your server. Verify that your collections can properly run? From the Server Home Page, select Help Review the documents included with each separately Center > User Resources. A set of installed collection to ensure that all requirements are documents that explains how to support met. In some instances, third-party software might be and use each collection is available on a required on the server or client machines to support a per-collection basis (organized in collection. If you have questions, contact alphabetical order). [email protected]. Share a protocol database across servers? Change the XMLDB Endpoint in the Clients running at multiple sites can access the same Admin Portal so that it points to the server protocols and components if they share the same that hosts the protocol database (Setup > XMLDB. Server Configuration). Authenticate users on the server? Configure authentication in the Admin You can restrict access to the server based on the Portal (Security > Authentication). identity of your network users. Pipeline Pilot supports File, Local, and Domain authentication. Enable impersonation? Enable impersonation in the Admin Portal Impersonation allows users to access network resources (Security > Authentication). using their network security credentials rather than their server account credentials.

Post-Installation | Page 27 Restrict publishing rights on the server? Configure access rights in the Admin Portal All users have read/write access to the shared tabs, (XMLDB > Access Rights). allowing all users the rights to “publish”. You can control how users publish protocols on these shared tabs by managing access rights. Support Web Port users? A link is available from the Server Home Any client user, who needs to run protocols remotely or Page for launching Web Port. from a web browser, needs to run Web Port. Provide database connectivity? Read the instructions in the Database ODBC drivers from DataDirect are included with the Integration Guide for the Integration Integration collection. These drivers support connections collection. ODBC Driver Help is also to Oracle, Microsoft SQLServer, MySQL, and DB2. A available on the Server Home Page. version of the Oracle JDBC driver for connecting to Oracle databases is also available. To use the Integration components with your database, you need to configure data sources on your server. Schedule protocol jobs? Install the Integration collection and run You can run protocols outside of the graphical client command-line program environment on any computer where Pipeline Pilot is RunProtocol.exe. Read the Command installed with job scheduling. Line Guide for the Integration collection. Migrate an existing Pipeline Pilot installation to a new Point all clients (Professional and web- machine? based) to the new host, and install the Reuse your protocol database (XMLDB) in the new required software on the new host. environment. Update your browser bookmarks to Web Port and any web pages that contain links to protocols. For further details, see Moving/Migrating an Installation. Upgrade an existing installation? Activate any non- Pipeline Pilot packages, Depending on your installed packages, some additional such as custom Java components or third- steps might be necessary to get your server properly party components, and recompile your configured and running. client applications that use the Java SDK or .NET SDK with the new versions of jalpp.jar or nalpp.dll. Upgrade an existing installation and modify the port Update your browser bookmarks to Web numbers? Port and any web pages that contain links Any applications that rely on the new port numbers will to protocols. need updating. Deploy Pipeline Pilot Server on a grid, cluster, web See the Pipeline Pilot Deployment Guide. farm, or reverse proxy? Ensure your machines are properly configured to support these features.

Page 28 | Pipeline Pilot Server • Server Installation Guide Configure the Java Server package that you installed as Before you can start the Java Server part of the server installation? services, validate the ports that you need Some additional steps might be necessary. to use in the Admin Portal (Processes > Server Ports). After validation, you can start the Java Server services from the Admin Portal. Add lines to your .bash_profile or .cshrc file to For the C-Shell, add the following lin to source either /linux_ ~/.cshrc bin/ppvars.sh or ppvars.csh. source/linux_ bin/ppvars.csh For the Bash shell, add the following line to ~/.bash_profile /linux_ bin/ppvars.sh where: the ~ in the path resolves to your home directory on the Linux server. is the root directory of the Pipeline Pilot installation.

Post-Installation | Page 29 Pipeline Pilot Server Home Page The Pipeline Pilot Home Page is a one-stop place for all things related to Pipeline Pilot, Pipeline Pilot Client, and other BIOVIA tools applications. It opens in a browser window and provides a way to get assistance and access documentation and programs. From here, you can install or upgrade Pipeline Pilot Client software, access documentation in the Help Center, and start Pipeline Pilot Web Port. Administrators can perform server tasks in the Admin Portal and install Data Source Builder and BIOVIA Draw.

Page 30 | Pipeline Pilot Server • Server Installation Guide Administration Portal System administration and server configuration tools are available in the Administration Portal (Admin Portal). This web application provides secure access from any network location, independent of the server location. It runs with Internet Explorer and Firefox or another Mozilla-supported browser. Tip: To access the Admin Portal, you need a valid user name and password combination. The default settings are: "scitegicadmin" (user name) and "scitegic" (password). The first time you log into the Admin Portal, you can use these credentials.

To log into the Admin Portal 1. From the Server Home Page, select Administration Portal. The login page opens. 2. Enter the default user name and password and click Log In.

The Admin Portal page Note: See the Admin Portal online help for further details on using this application.

Post-Installation | Page 31 Chapter 7: Deployment Options

Load Balancing Deployments Load Balancing Features Load balancing distributes incoming HTTP requests across web servers in a server farm, to avoid overloading any one server. Because load balancing distributes the requests based on the actual load at each server, it is excellent for ensuring availability and defending against denial of service attacks. This is meant to be a scalable system for supporting high levels of web-based requests for fast operations, and its significant restrictions make it unsuitable for general use. IMPORTANT! If you have an enterprise deployment and have a BIOVIA Foundation Hub server, follow the instructions in the BIOVIA Foundation Hub Deployment Guide for configuring load balancing. Load balancing is supported on both Windows and Linux. Due to the differences in paths, only one type of operating system can be deployed across the enterprise. Pre-configuration Steps Before you begin, carefully review the Pipeline Pilot Admin Portal Guide sections Configuring a Reverse Proxy and Load Balancing Deployments. Also see the Web Farm section of the Pipeline Pilot Deployment Guide for a higher level overview of Load Balanced Server environments. Be sure to have the following: The server names for each server. For example: server1.domain.local server2.domain.local server3.domain.local ... There should also be a load balancer server, for example: onelab..local To further configure these servers for BIOVIA Foundation Hub, see the BIOVIA Foundation Hub 2016 Deployment Guide. Configuration Steps Configure a Reverse Proxy 1. Navigate to the Admin Portal and go to Setup > Reverse Proxy and Load Balancing. 2. Set the Reverse Proxy Name and Aliases to point to your chosen Reverse Proxy address. 3. Check the Load Balanced box. Configure each Load Balanced Server 1. Install Pipeline Pilot Server 2016 on each machine and in the same location.

Page 32 | Pipeline Pilot Server • Server Installation Guide 2. Create a Shared User Directory, Jobs Directory, and Shared Public Directory. The XMLDB is not shared and exists separately on each deployed server. It will be read only and locked when you select the Load Balanced box above. 3. Each node requires identical settings for various other settings such as security, group and permission settings, proxy definition, etc. a. Create your settings on Server 1. This will be set. b. Go to Maintenance > Export Import Configuration and export the configuration. c. On each additional server, navigate to Maintenance > Export Import Configuration and import that configuration file . IMPORTANT! In Windows: Run LoadBalancedConfig_Enable.reg on each node. This will add information to the Registry. You can find the file in /templates/windows_noarch.

Tip: Check the settings in the admin portal to ensure that the shared folders and reverse proxy and load balancing settings are correctly applied. If they are not they can be set manually or enacted by copying the file /xmldb/Objects/Registry.2.xml into the equivalent location on the subsequent servers. Test load balancing by stopping Pipeline Pilot Services on all but one server and verifying that access via the load balanced address functions. Repeat for each server. Restart the Pipeline Pilot Services when you are finished.

Deployment Options | Page 33 Grid Engine Installation A grid installation is virtually identical to a cluster installation. The only real difference is that instead of running Pipeline Pilot on the cluster nodes, the grid engine software manages the cluster nodes. For more information, see the Pipeline Pilot Deployment Guide. Grid Engine setup When using a grid engine, ensure that you can submit jobs from the host where you are installing Pipeline Pilot. The server does not have to be installed on the head node of the grid, but it must be installed on a node that can submit jobs to the grid. A best practice is to use the edge node configuration. Because the server is not a member of the grid, it can submit jobs to the grid using this configuration. Grid Engine requirements See the separate document Pipeline Pilot System Requirements for supported grid engines.

Notes: Install Pipeline Pilot on a Linux system that can support your distributed grid computing system. There are no additional hardware requirements beyond what is specified for a normal installation (see the Pipeline Pilot System Requirements document). The Admin Portal includes grid engine support settings, including type, path, and default queue name. It also provides a list of queue names for a client to use at run time, but this does not affect the operation of the server in any way. Pipeline Pilot uses command-line options that are generic for the NQS commands supported by the grid engines. Custom grid engines are currently not supported. To view and create community topics on unsupported grid engines, visit the public discussion area of the BIOVIA Community at https://community.3dsbiovia.com/.

Setting the User Account for Running Xvfb on Grid Cluster Nodes If you did not run the scirootinstall script when you installed Pipeline Pilot on the primary node, edit the scixvfb script in the Pipeline Pilot linux_bin directory, and set the user account for running xvfb on the grid cluster nodes. Tip: Xvfb is required by components that use R software to generate plots and graphics. For further details, see the R Software Support Guide. Look for the line: USER= Change it to: USER=ppuser Where ppuser is the user account specified when you installed Pipeline Pilot. Note: If you already ran scirootinstall, this step is not required.

Page 34 | Pipeline Pilot Server • Server Installation Guide

Chapter 8: Uninstalling Pipeline Pilot

IMPORTANT! Before uninstalling software on any server, do the following (in order): 1. Shut down all web services that are using the ports for the software you are about to remove. See Pipeline Pilot Services. 2. Confirm there are no jobs running on the server. In the Admin Portal, go to Jobs > Running Jobs. 3. Back up your XMLDB. In the Admin Portal, go to XMLDB > Backup. 4. If you customized the location of your XMLDB directory, reconfigure it to the /xmldb default location. In the Admin Portal, go to Setup > Folder Locations. 5. If the server is connected to a BIOVIA Foundation Hub server, disconnect it from the Security > Hub Connection page in the admin portal before uninstalling the server.

Uninstalling from a Windows Server Use the Windows Services console to shut down all services that are currently using the ports for the software you need to remove. For more information, see Pipeline Pilot Services. (You can also stop all Pipeline Pilot services when you start the Pipeline Pilot installer, but you still have to manually shut down Apache for earlier releases.) 1. Open the Windows Services console. 2. Shut down BIOVIA Pipeline Pilot 16.1.0 (Manager). Note that shutting down this service will also shut down the following sub-services: BIOVIA Pipeline Pilot 16.1.0 Service (Httpd) and BIOVIA Pipeline Pilot 16.1.0 Service (Tomcat). 3. Use Add/Remove Programs in the Windows Control Panel to uninstall the server software. For Windows 8.1 or above, right-click the Pipeline Pilot icon in the Windows Start Menu and select Uninstall Tip: When you uninstall Pipeline Pilot, your protocol database (XMLDB) remains on your server. You can install a new version in the same location. You can also migrate your XMLDB to a new server if you plan to reuse it somewhere else. See Moving/Migrating an Installation.

Uninstalling Pipeline Pilot Server | Page 36 Uninstall from a Linux Server 1. Run ppvars.sh before stopping the server. 2. Stop the server using the stopserver command in the /opt/BIOVIA/installdir/linux_bin directory. ./stopserver 3. Remove the boot startup and shutdown scripts: rm /etc/rc.d/rc*.d/*ppilotd. where is the server port number rm /etc/init.d/ppilotd. where is the server port number 4. If you installed on a cluster or a grid, for each node in the cluster: Stop the server (cluster only): /opt/BIOVIA/installdir/linux_bin/stopserver Unmount the exported file system: umount -f /opt/BIOVIA Edit /etc/fstab and remove the entry for the exported file system. Remove the scitegic mutex directory (cluster only): rm -rf /var/tmp/scitegic-mutexes- where is the server port number Remove the /etc/pam.d/scitegic file (cluster only): rm /etc/pam.d/scitegic If you installed the boot scripts: Remove the startup and shutdown scripts, see Step 3 (cluster only). If you installed the SELinux module, remove it: /usr/sbin/semodule -r scitegic If you installed the Xvfb init script (grid only), remove it: rm /etc/rc.d/rc*.d/*scixvfb. where is the server port number rm /etc/init.d/scixvfb. where is the server port number 5. If you installed on a cluster, remove the export entries: Edit /etc/exports and remove the entries for the exported file systems. 6. Remove the scitegic mutex directory: rm -rf /var/tmp/scitegic-mutexes- where is the server port number 7. Remove the installation: rm -rf /opt/BIOVIA/installdir 8. Remove the user and group entries for the Pipeline Pilot Server: userdel -r

Page 37 | Pipeline Pilot Server • Server Installation Guide where is the name of the Apache user. groupdel where is the name of the Apache group. 9. Remove the /etc/pam.d/scitegic file: rm /etc/pam.d/scitegic 10. If you installed the SELinux module, remove it: /usr/sbin/semodule -r scitegic

Uninstalling Pipeline Pilot Server | Page 38 Appendix A: Unattended Installation on Linux

Note: The following example details an unattended Linux installation, but the procedure also applies to a Server installation.

Setting Required Pre-Installation Environment Variables The following parameters must be set before you begin an installation, an upgrade, a reinstall, or an add/remove packages begins. IMPORTANT! You must be running in the bash shell. ROOT=/tmp/pps (your installation directory) export ROOT

SCITEGIC="$ROOT" export SCITEGIC

PACKAGEDIR=apps/scitegic/core/packages_linux64 export PACKAGEDIR

ODBCINST="$ROOT/$PACKAGEDIR/datadirect/odbcinst.ini" (ODBC driver installation environment) export ODBCINST

MEDIADIR="/tmp/mydir" (path to the installer source that contains the bin directory) Tip: Navigate to the directory where the bin directory is locate, enter pwd, and then copy that value to the MEDIADIR variable.

Note: All the directories/files in this path must have execute permissions. export=MEDIADIR

LD_LIBRARY_PATH="$ROOT/linux_ bin":"$MEDIADIR":"$MEDIADIR/bin":"$ROOT/$PACKAGEDIR/datadirect/lib": "$ROOT/apps/scitegic/core/packages_linux64/openssl/openssl- 0.9.8o/lib":"$ROOT/$PACKAGEDIR/openssl/openssl- 0.9.8j/lib":"$ROOT/$PACKAGEDIR/openssl/openssl- 0.9.8e/lib":"$ROOT/$PACKAGEDIR/openssl/openssl- 0.9.8a/lib":"$ROOT/$PACKAGEDIR/openssl/openssl-0.9.8u/lib" export LD_LIBRARY_PATH

PERL5LIB="$ROOT/$PACKAGEDIR//perl-5.8.9" export PERL5LIB

Unattended Installation on Linux | Page 39 SCITEGICMUTEXES=/var/tmp/scitegic-mutexes/9943-myuser Note: The scitegic-mutexes directory must have 777 permissions assigned. export SCITEGICMUTEXES

Installation Environment Variables To run the installer for a new, first-time installation the following command is entered: "$MEDIADIR/bin/installcmdline" -execute -d "$ROOT" -l "$LICENSEPATH" -p $HTTPPORT -s $HTTPSPORT -c "$CLUSTERMETHOD" -m "$SCITEGICMUTEXES" -mode $INSTALLMODE Note: For information on MEDIADIR and $ROOT, see the pre-installation parameters.

Where: LICENSEPATH = For example: /tmp/license HTTPPORT = -p DEFAULT_HTTPPORT For example: -p 9944 HTTPSPORT = -s DEFAULT_HTTPSPORT For example: -s 9943

CLUSTERMETHOD = -c

The options are: -c round-robin -c job_leveling no -c option: no clustering INSTALLMODE = -mode "FirstTimeInstall" Other options are: Upgrade Reinstall AddRemove

Package names.

Notes: The following packages include collections that may not be installed with the regular Pipeline Pilot license (e.g, scitegic/discoverant, scitegic/thomsonreuters, scitegic/scnotebook, scitegic/scitegic/scmobile, scitegic/notebook). This list does not include packages from Applications (e.g., Insight, Request, Materials Studio, Discovery Studio, etc.)

Page 40 | Pipeline Pilot Server 2016 • Server Installation Guide scitegic/admetox scitegic/admetoxdata scitegic/admin scitegic/adminportal scitegic/analyticalinstrumentation scitegic/appcatalog scitegic/bioutilities scitegic/cem scitegic/chemistry scitegic/chemistryadv scitegic/chemistrycart scitegic/chemistrycontent scitegic/chemistrydata scitegic/chemistrysdk scitegic/chemistrytlkt scitegic/chemmining scitegic/clientsdk scitegic/clientsdkdotnet scitegic/clientsdkjava scitegic/clientsdkjavascript scitegic/core scitegic/coreutils scitegic/dataaccess scitegic/discoverant scitegic/docs scitegic/dsbuilder scitegic/gems scitegic/gemsapp scitegic/geneexpression scitegic/geneexpressiondata scitegic/generic scitegic/hub scitegic/imaging scitegic/imagingadv1 scitegic/imagingdata scitegic/installer scitegic/integration scitegic/javaserver scitegic/labelprinting scitegic/launcher scitegic/listmgmt scitegic/massspecprot scitegic/massspecprotdata

Unattended Installation on Linux | Page 41 scitegic/modeling scitegic/modelingadv scitegic/mongodb scitegic/plateanalytics scitegic/platform scitegic/polymerproperties scitegic/ppclient scitegic/queryservice scitegic/report scitegic/reportadv scitegic/rss scitegic/rstats scitegic/scheduler scitegic/scmobile scitegic/scnotebook scitegic/seqanal scitegic/seqanaldata scitegic/sharepoint scitegic/systemlibs scitegic/textanalytics scitegic/thomsonreutersanalytics scitegic/thomsonreutersclinical scitegic/thomsonreuterscommon scitegic/thomsonreutersdrugdesign scitegic/thomsonreutersdrugs scitegic/thomsonreutersontologies scitegic/thomsonreuterspatents scitegic/thomsonreuterstargets scitegic/waf scitegic/web scitegic/websvc

Installation Command Example

Note: A space is required between each package name when installing multiple packages at the same time. To run the installer for a new, first-time installation the following command is entered. This example installs the default Platform. "$MEDIADIR/bin/installcmdline" -execute -d $ROOT -l /tmp/license -p 9944 -s 9943 -c round-robin -m SCITEGICMUTEXES -mode FirstTimeInstall -pi scitegic/admin -pi scitegic/adminportal -pi scitegic/appcatlog -pi scitegic/clientsdk -pi scitegic/clientsdkdotnet -pi scitegic/clientsdkjava - pi scitegic/clientsdkjavascript -pi scitegic/core -pi scitegic/coreutils -

Page 42 | Pipeline Pilot Server 2016 • Server Installation Guide pi scitegic/dataaccess -pi scitegic/docs -pi scitegic/dsbuilder -pi scitegic/gems -pi scitegic/gemsapp -pi scitegic/generic -pi scitegic/integration -pi scitegic/javaserver -pi scitegic/launcher -pi scitegic/mongodb -pi scitegic/queryservice -pi scitegic/report -pi scitegic/reportadv -pi scitegic/rss -pi scitegic/scheduler -pi scitegic/sharepoint -pi scitegic/waf -pi scitegic/web -pi scitegic/websvc

Unattended Installation on Linux | Page 43 Appendix B: Configuring Apache as a Load Balancer and Reverse Proxy There are no known requirements for the specific load balancer hardware or software. The following information is intended to guide you through the process of configuring the free Apache server to create a load balanced configuration without any additional hardware requirements. To enable an external Apache server to act as a load balancer, it is necessary to create or modify a few files. If you are using RedHat 4, update your Apache installation, as these features are relatively new. (RedHat 5 and later already has the required features.) To load balance or reverse proxy HTTP requests, you need a proxy_balancer.conf file, (by default, this file is available in /etc/httpd/conf.d). To handle SSL requests, you need an ssl.conf file, (by default, this file is available in /etc/httpd/conf.d). Note: For all load balanced and reverse proxy configurations, ProxyPreserveHost must be enabled.

HTTP Requests proxy_balancer.conf #LoadModule proxy_balancer_module modules/mod_proxy_balancer.so

# ServerAdmin [email protected] # DocumentRoot /www/docs/dummy-host.example.com # ServerName dummy-host.example.com # ErrorLog logs/dummy-host.example.com-error_log # CustomLog logs/dummy-host.example.com-access_log common ProxyPass / balancer://mycluster/ ProxyPreserveHost On BalancerMember http://node1:9944 BalancerMember http://node2:9944 BalancerMember http://node3:9944 #the following line allows connections to the balancer from any client IP address #you may choose to restrict access to specific subnets or domain names as you wish Allow from all

Configuring Apache as a Load Balancer and Reverse Proxy | Page 44 SSL Requests To connect to the server via SSL: The best place to put this is at the end of the VirtualHost section in your ssl.conf file. To connect to the Pipeline Pilot Server via SSL, turn on the SSLProxyEngine. ssl.conf

# SSL Proxy Load Balancer SSLProxyEngine Off ProxyPass / balancer://mycluster/ ProxyPreserveHost On #by proxying to the http addresses, we can in essence provide SSL Termination at the proxy server #BalancerMember https://host1:9943 #BalancerMember https://host2:9943 #BalancerMember https://host3:9943 BalancerMember http://host1:9944 BalancerMember http://host2:9944 BalancerMember http://host3:9944 #the following line allows connections to the balancer from any client IP address #you may choose to restrict access to specific subnets or domain names as you wish Allow from all

Page 45 | Pipeline Pilot Server 2016 • Server Installation Guide Appendix C: Installing Separately Licensed Collections The Thomson Reuters and Materials Studio collections require a separate license file. Download the appropriate license file from the BIOVIA Community website at https://community.3dsbiovia.com/. PPTRCollection.lic PP_MSC_2018.lic Note: The Materials Studio Collection (MSC) license can be found in the Materials Studio/Materials Science section of the BIOVIA Download Center. The Notebook license can be found in the Pipeline Pilot section.

Materials Studio Collection The MSC components are licensed using the License Pack and the MSC requires the latest version of the License Pack. This can be obtained from the Materials Studio/Materials Science section of the BIOVIA Download Center. Note: If you have already installed Materials Studio 2016 on the machine you are using as your Pipeline Pilot server you do not need to install the License Pack again. 1. Go to the BIOVIA Community website at https://community.3dsbiovia.com/.. 2. Download the correct version of Materials Studio for your platform. 3. Run the EXE or TGZ file to decompress the installable files. 4. Locate and run lp_setup.exe (in the LicensePack folder) to install the License Pack. IMPORTANT! After installation of the License Pack it should be configured according to the instructions in the "Licensing Materials Studio" chapter of the BIOVIA Materials Studio 2016 Installation and Administration Guide which is available on the BIOVIA Download center in the same location as the License Pack.

Windows (upgrade) 1. Copy the .lic file to this location: \licensing\Licenses. 2. Run the server installer. 3. When prompted , choose Install to an existing location. Upgrade any existing protocols. 4. When prompted to confirm your license, click Next. Do not attempt to browse to a license file. 5. When prompted to Select products to install, ensure that the appropriate collection is selected. 6. Complete the remaining installer dialogs.

Windows (new installation) 1. Run the server installer to install Pipeline Pilot 2016. 2. Copy the .lic file to this location: \licensing\Licenses. 3. Re-run the server installer. 4. When prompted, choose Add/Remove Products. 5. When prompted to confirm your license, click Next. Do not attempt to browse to a license file.

Appendix C: Installing Separately Licensed Collections | Page 46 6. When prompted to Select products to install, ensure that the appropriate collection is selected. 7. Complete the remaining installer dialogs.

Linux (upgrade) 1. Copy the .lic file to this location: /licensing/Licenses 2. Run the server installer. 3. When prompted, choose 3. Upgrade an existing installation. 4. Enter the location of your Pipeline Pilot Server. 5. When prompted for a license, press Enter. Do not attempt to browse to a license file. 6. Select the default action when prompted to install the selected collection(s) by pressing oEnter. 7. Accept the remaining defaults. 8. For the MSC, open the Pipeline Pilot Admin Portal and select the Setup > Global Properties page. Set the LicensePackDir property to the directory where the BIOVIA License Pack is installed. Click the Update button.

Linux (new installation) 1. Run the server installer to install Pipeline Pilot 2016. 2. Copy the .lic file to this location: /licensing/Licenses 3. Re-run the server installer. 4. When prompted, choose 5. Add/Remove packages from an existing installation. 5. Enter the location of your newly installed Pipeline Pilot Server. 6. When prompted for a license, press Enter. Do not attempt to browse to a license file. 7. Select the default action when prompted to install the selected collection(s) by pressing Enter. 8. Accept the remaining defaults. 9. For the MSC, open the Pipeline Pilot Admin Portal and select the Setup > Global Properties page. Set the LicensePackDir property to the directory where the BIOVIA License Pack is installed. Click Update.

Page 47 | Pipeline Pilot Server 2016 • Server Installation Guide Appendix D: Required Linux RPMs

Red Hat Enterprise Linux 7 alsa-lib (i686) audit-libs (i686) compat-libstdc++-33 (i686) e2fsprogs-libs (i686) expat (i686) glibc (i686) keyutils-libs (i686) krb5-libs (i686) libgcc (i686) libICE (i686) libpcap (i686) libselinux (i686) libsepol (i686) libSM (i686) libstdc++ (i686) libX11 (i686) libXau (i686) libXdmcp (i686) libXext (i686) libXi (i686) libXmu (i686) libXpm (i686) libXpm (x86_64) libXrender (i686) libXt (i686) libXtst (i686) pam (i686) cairo (i686) fontconfig (i686) glib2 (i686) bzip2-libs (i686) freetype (i686) libjpeg-turbo (i686) libtiff (i686) zlib (i686) compat-libtermcap (i686) libdrm (i686)

Appendix D: Required Linux RPMs | Page 48 libXdamage (i686) libXfixes (i686) libXxf86vm (i686) libXinerama (i686) tcsh bc

Red Hat Enterprise Linux 6 audit-libs (x86_64) bzip2-libs (x86_64) cairo (x86_64) compat-libstdc++-296 (iX86) compat-libstdc++-33 (iX86) cracklib (iX86) cyrus-sasl-lib (x86_64) db4 (iX86) e2fsprogs-libs (x86_64) expat (iX86) expat (x86_64) fontconfig (iX86) fontconfig (x86_64) freetype (iX86) freetype (x86_64) gamin (iX86) glib2 (iX86) glib2 (x86_64) libdbi (iX86) libdrm (iX86) libdrm (x86_64) libICE (iX86) libjpeg (x86_64) libselinux (iX86) libSM (iX86) libstdc++ (iX86) libtiff (x86_64) libuuid (iX86) libX11 (iX86) libXau (iX86) libxcb (iX86) libXcursor (iX86) libXdamage (iX86) libXdamage (x86_64)

Page 49 | Pipeline Pilot Server 2016 • Server Installation Guide libXdmcp (iX86) libXext (iX86) libXfixes (iX86) libXfixes (x86_64) libXi (iX86) libXinerama (iX86) libXinerama (x86_64) libXp (iX86) libXp (x86_64) libXpm (iX86) libXrandr (iX86) libXrender (iX86) libXt (iX86) libXtst (iX86) libXxf86vm (iX86) libXxf86vm (x86_64) mesa-dri-drivers (iX86) mesa-libGL (iX86) mesa-libGL (x86_64) mesa-libGLU (iX86) mesa-libGLU (x86_64) ncurses-libs (x86_64) nspr (iX86) pam (iX86) pixman (x86_64) zlib (iX86)

SUSE Linux Enterprise 11 audit-libs-32bit (x86_64) cairo (x86_64) fontconfig (x86_64) fontconfig-32bit (x86_64) freetype2 (x86_64) freetype2-32bit (x86_64) keyutils-libs-32bit (x86_64) krb5-32bit (x86_64) libbz2 (x86_64) libbz2-1-32bit (x86_64) libcom_err2-32bit (x86_64) libdrm (x86_64) libdrm-32bit (x86_64) libexpat1 (x86_64)

Appendix D: Required Linux RPMs | Page 50 libexpat1-32bit (x86_64) libglib-2_0-0 (x86_64) libglib-2_0-0-32bit (x86_64) libgthread-2_0-0 (x86_64) libgthread-2_0-0-32bit (x86_64) libjpeg (x86_64) libncurses5-32bit (x86_64) libncurses5 (x86_64) libpixman (x86_64) libstdc++46-32bit (x86_64) libtiff3 (x86_64) libuuid1-32bit (x86_64) Mesa-32bit (x86_64) Mesa (x86_64) pam-32bit (x86_64) pcre-32bit (x86_64) pcre (x86_64) xorg-x11-libICE-32bit (x86_64) xorg-x11-libs-32bit (x86_64) xorg-x11-libSM-32bit (x86_64) xorg-x11-libX11-32bit (x86_64) xorg-x11-libXau-32bit (x86_64) xorg-x11-libxcb-32bit (x86_64) xorg-x11-libXext-32bit (x86_64) xorg-x11-libXfixes-32bit (x86_64) xorg-x11-libXfixes (x86_64) xorg-x11-libXpm-32bit (x86_64) xorg-x11-libXrender-32bit (x86_64)

Page 51 | Pipeline Pilot Server 2016 • Server Installation Guide