| Oracle® Health Sciences Data Management Workbench Installation Guide Release 2.4 E52291-03 |
|
|
PDF · Mobi · ePub |
Home > Installation Guide > Configure File Watcher Support
| Oracle® Health Sciences Data Management Workbench Installation Guide Release 2.4 E52291-03 |
|
|
PDF · Mobi · ePub |
Home > Installation Guide > Configure File Watcher Support
The File Watcher utility checks a file system location that you specify for data files whose name matches a pattern you specify and loads them into DMW for use in a particular study.
File Watcher can detect and load two types of files:
SASâincluding CPort or XPort formats, a single dataset, or a single dataset in a .zip file.
Textâincluding .txt or .csv files that may be contained in a .zip file.
This section contains the following topics:
The Oracle LSH Distributed Processing (DP) Server is required for File Watcher. Following instructions in the Oracle Life Sciences Data Hub Installation Guide, you have already set up the Distributed Processing (DP) Server, including:
Installed the DP Server where it has access to Oracle SQL*Loader. It must also have access to the directories where you will put text data files to be loaded into Oracle DMW; see Section 6.2, "Create Directories".
Installed the DP Server where it has access to the SAS processing engine and the directories where you will put SAS data files to be loaded into Oracle DMW.
If the same DP Server has the access required for both SAS and text files, you need only one DP Server.
Started the DP Server with the File Watcher service enabled. The File Watcher service detects files to be loaded.
Defined a Service Location in the Oracle LSH user interface for each computer where the DP Server is installed.
Note:
The system populates the list of values in Oracle DMW's Watcher Listing for the Text and SAS DP Servers with the Service Locations you define.Defined a Text for SQL*Loader service on the Service Location that has access to Oracle SQL*Loader and the text files to be loaded. This service loads data from text files into the Oracle DMW database.
Defined a SAS service on the Service Location that has access to SAS and the SAS files to be loaded. This service loads data from SAS files into the Oracle DMW database.
See the Oracle Life Sciences Data Hub System Administrator's Guide for more information on Service Locations and services.
You must create the directories that will serve as the Watched Folders that File Watcher monitors for data files. You can choose to archive loaded data files; in which case you must create directories in which to place the archived files.
Create nested directories on one or more computers to which the DP Server has access:
A root folder on each computer from which files are loaded into Oracle DMW.
Subfolders
Six folders: one for each combination of the two file types (SAS and text) and the three lifecycle areas (Development, Quality Control, and Production). These folders can be on different computers, but there can be only six across the installation. Oracle recommends using a naming convention that includes the file type and lifecycle mode.
Alternatively, you can use just three folders, one for each lifecycle area and both file types.
A DP Server must have access to each of these folders; for example, you can set up an NSF mount of the file system to each computer where a DP Server is installed.
All studies in this Oracle DMW instance must use the same three or six root folders for their input data files. The system creates a study-specific subfolder in each root folder using the name you specify; instructions are in the Administration chapter of the Oracle Health Sciences Data Management Workbench User's Guide. The study-specific subfolders become the watched locations for the study.
You must enter the location of these folders as system profile values; see "Registering Folder Locations" in the Administration chapter of the Oracle Health Sciences Data Management Workbench User's Guide.
Note:
If the study will use only SAS files or only Text files for data loading, you do not need to create subfolders for the other file type. If you do not plan to use the Quality Control/UAT lifecycle area, you do not need to create subfolders for that lifecycle area.If you want to archive data files after their data has been loaded, create directories in which to archive them. They must have the same structure as the watched foldersâeither six or three folders. Oracle recommends creating them on the same UNIX file system as the watched folders.
On the computers where labs or other data sources will post data files, restrict access to data files to prevent investigators and others from seeing data they should not see, such as blinded data and data from other sources.
Set up a secure file transfer method:
Transfer files using a secure system process such as SFTP or a secure file synchronization service.
Isolate the file share from direct user access by using an indirect method of conveying files to the file share such as a dropbox.
Use UNIX/Linux file system security by creating a UNIX/Linux file system group and assigning user IDs that require access to the group:
The OS user ID that executes the DP Server that is running File Watcher. You set up this account during Oracle LSH installation.
OS user IDs that place files in the file share (using SFTP, for example).
The OS user ID that executes a program that places files in the file share (an automated file transfer or file synchronization program).
Restrict access to the file share. Only the file owner and group should have any permissions. The other permissions should be empty. If the user ID running the DP Server is the same as the user ID used to transfer all files, then only owner file permissions are needed. Assuming this is not the case, grant permissions to both the file owner and group as follows.
In /etc/bashrc or /etc/profile, set umask:
umask 007
This has the same effect as the following settings:
For File Watcher to load the data files it detects in a timely manner, the date and time settings on your serversâdatabase server, the middle tier server where the WebLogic Server is installed, and the middle tier server(s) where the Oracle LSH Distributed Processing (DP) Server is installedâmust all have the same date and time setting.
If your servers are located in different time zones, do one of the following:
To determine valid values for the TZ environment variable, see Section 6.4.3, "Use the tzselect Utility".
Finally, follow Section 6.4.4, "Verify the Time Zone Setting Used by Java Virtual Machine (JVM)".
You can set the TZ environment variable as follows:
TZ='<correct timezone>'export TZ
For example, to set the time zone to the US Pacific time zone:
TZ='America/Los_Angeles'export TZ
To determine a valid TZ value, see "Use the tzselect Utility".
You can set the TZ environment variable automatically by one of the following methods:
Set the TZ variable in the startup configuration file (.profile or .cshrc file) for the operating system user that runs the DP Server process or the Web Logic Server process on the respective middle tier.
Set the TZ variable in the shell script that starts the process:
For the DP Server, set the TZ variable in the DP Server Home/cdr_apps_dpserver.sh shell script, or in a shell script that invokes cdr_apps_dpserver.sh; see the Oracle Life Sciences Data Hub Installation Guide for information
For the Web Logic Server, set the TZ variable in the setDomainEnv.sh script from your Web Logic installation in a full path similar to middleware_home/user_projects/domains/base_domain/bin/setDomainEnv.sh.
To determine a valid TZ value, see "Use the tzselect Utility".
Many UNIX systems provide the tzselect utility to aid in selecting a Posix standard time zone format. You enter this command at the UNIX command prompt with no arguments and it prompts you to select a continent or ocean and then a time zone region. It returns the string to use in the TZ command.
See your operating system documentation for more information about this command in your specific environment.
On the middle tier servers DMW is implemented using Java, and you must confirm that the Java Virtual Machine (JVM) is using the correct time zone
Note that a correct return from the UNIX date command does NOT mean that the Java Virtual Machine (JVM) is using the correct time zone. The JVM looks for the time zone settings as follows. This can vary in different implementations of UNIX/Linux, so check your operating system documentation for further information:
JVM uses the environment variable TZ if it is set.
If TZ is not set, then JVM looks for the file /etc/sysconfig/clock and finds the ZONE entry.
If neither TZ nor ZONE is set, JVM compares the contents of /etc/localtime to the files in /usr/share/zoneinfo looking for a match. The matching path and filename under /usr/share/zoneinfo provides the time zone.
Check the time zone that the DP Server and Web Logic Server are actually using by reading the time zone information displayed at the beginning of most lines in their log files:
DP Server: The log file is located in the directory: DP Server Home/log.
Web Logic Server: The log file is located in a full path similar to: /app/product/middleware/user_projects/domains/base_domain/servers/DMWServer/logs.
In addition, for each data source in each study a File-type input clinical data model must be defined; see the clinical data model chapter of the Oracle Health Sciences Data Management Workbench User's Guide.
On the Administration page of Oracle DMW you can monitor study File Watchers and start and stop them.
|
 Copyright © 2013, 2018, Oracle and/or its affiliates. All rights reserved. Legal Notices |
|