Knowledge Base

  

The following articles represent a subset of our searchable helpdesk database. Please use your browser's find facility and the links below to locate articles.

Quick Links 

All Products 

Title Versions known to be affected Article
Compiler Installation 9.0.x

A supported Compiler must be installed before you begin the installation for Prophet Professional.

We recommend using the latest version of Visual Studio that is supported by your Prophet and Operating System wherever possible.

Visit the following link for further details and download options for supported Compilers:

 

There is a known issue where Microsoft Updates can break the SDK (7.x) version of the Compilers and Microsoft will no longer provide fixes to the SDK compiler. We therefore recommend that you use a Visual Studio compiler to ensure the best option for ongoing support from Microsoft.

 

Visual Studio Express requires product registration with Microsoft. Each user needs to create a Microsoft account and activate their product against their account. This is part of the terms of use of the Microsoft Visual Studio Express product. Other versions of the Visual Studio products are available and may be licensed by your organisation.




Return to Top

DCS Items 

Title Versions known to be affected Article
Compiler Installation 9.0 / 9.0.1

A compiler is required to run DCS programs and needs to be installed prior to DCS. Supported compilers are available from Microsoft via the Visual Studio suite.

Compiler support

The following compilers are supported in DCS 9.0:

• Microsoft Visual Studio 2012 (including Express Edition)

• Microsoft Visual Studio 2013 (including Express Edition)

 

More Information about Supported Compilers are listed on Prophet Web on Supported Compilers page.

 

Selecting the compiler in DCS

Open the DCS application and navigate to the File > Options menu.

Select your compiler and build architecture from the drop-down boxes provided.

Compiler Issues after installing .Net 4.5.1
9.0 / 9.0.1

Upgrading .Net to version 4.5.1 will break older compilers and cause errors during the compile stage. To view errors generated by the compiler open the produced .MSG file (located in your DCS project folder) in a text editor such as notepad. Any error codes and messages will appear towards the end of the file.

 

We recommend that you install and use the latest Visual Studio compilers to avoid further issues with the older SDK compilers.

 

Compiler Error: "LNK1123: failure during conversion to COFF: file invalid or corrupt".

There are several methods to resolve this error depending on which compiler you have installed:

•Users with the Visual Studio 2010 compiler should upgrade to VS 2010 SP1

•Users with the Visual Studio 2012 compiler should upgrade to VS 2012.4

•Users with Windows SDK 7.1 should install the following patch:

http://www.microsoft.com/en-gb/download/details.aspx?id=4422

 

Further details on this issue can be found via the Microsoft Support site as below:

http://support.microsoft.com/kb/2757355

Note that applying the above patch to the SDK can expose additional errors with the compiler (see below).

 

Compiler Error: “C1083: Cannot open include file: 'ammintrin.h': No such file or directory”

This is caused by Microsoft updates that have broken the SDK and 2010 compiler libraries. Microsoft have stated that they do not intend to support or fix this bug:

https://connect.microsoft.com/VisualStudio/feedback/details/660584

 

To resolve this issue you will need to obtain a copy of the missing files from the Visual Studio 2012 or 2013 suite (including the Express editions). Since you need to install a version of Visual Studio to resolve this, many clients have opted to simply use the new Visual Studio compiler instead of fixing the old SDK/2010 compilers.

 

Install Visual Studio 2012/2013 and manually copy the missing ammintrin.h file from the new Visual Studio folder into the "Visual Studio 10.0" folder for the SDK/2010 compiler.

For example, you should copy from VS11:

c:\Program Files (x86)\Microsoft Visual Studio 11.0\vc\include\ammintrin.h

into VS10:

c:\Program Files (x86)\Microsoft Visual Studio 10.0\vc\include\ammintrin.h

Date types for input formats  All

When creating an Input Format for DCS that uses a date field you can click on the Date Format cell to bring up a new window that allows you to define the formatting used.

The Date Format drop-down box offers a list of commonly used date formats, where you can also directly type any new formats that you use. For example, the date type mdyyyy (or m/d/yyyy) does not appear on the default list but can be manually entered.

Note that if you require a separator character that this is not displayed in the drop-down but is included as an additional option below the Date Format in the Separator section.

For the above example, if you wish to use the Date Format m/d/yyyy then you need to enter mdyyyy as the Date Format and select the / character in the Separator section.

 

When creating an Output Format that contains a date field please note that the format used is taken from the current user’s Regional Settings. This means that different users can run the same program but produce output tables that have different date formats.

If you supply output files from one DCS program to be the input files for another program, and you use different date formats within your organisation, then be aware that you may need to regularly change the input format settings to read files generated by different users. A solution to this is to convert your output date variables into text fields that are configured using a standard date format.

Using the above example, you can convert your output date from any Regional Settings format into a text variable using the following code:

MyText = Month(MyDate) + "/" + Day(MyDate) + "/" + Year (MyDate)

This can be read into the next DCS program using the expected date format of m/d/yyyy.

Grouping and Sorting memory errors All

Grouping and Sorting can use large amounts of memory in DCS. The best practice is to Limit DCS memory usage.

Generally, a memory setting of 100 MB in most cases provides good performance and should allow output files from large input files to be sorted.

 

When the specified amount of memory is fully used DCS writes the output lines currently in memory to temporary files in a TEMP subdirectory of the output file location, clears the output lines from memory and then continues processing the input file. When DCS has finished processing the entire input file it reads the temporary output files in parallel and creates the final output file. All the temporary files related to an output file need to be open simultaneously while DCS creates the final sorted, grouped or summarized output file.

For example, if you have a 100 MB input file and have specified 12 MB of memory for sorting, grouping and summarizing, DCS will create approximately 9 temporary files in the TEMP directory for each final output file.

 

As a guide, if you want to specify the amount of memory that DCS uses for sorting, grouping and summarizing then it should be set to at least 1% of the size of the input file. Therefore, for a 2000 MB input file you should use at least 20 MB of memory.

Use the Options to limit the memory to below 2GB to improve performance.

Intex functions requirements All

The Intex database libraries are licenced directly from Intex Solutions Inc. and are not provided by FIS.

 

FIS provides a licenced plug-in that allows you to access Intex functions from within the Prophet and DCS applications.

Once the FIS plug-ins have been installed you need to select the DCS Options to enable the Intex version available, located under the DCS»Options»Third Party Options menu.

 

You will also need to set the CDI and CDU folder paths within the Run Settings for each job, located under the Workspace»Run Settings»Intex section of the run settings.

Prophet generated files contain unreadable characters All

To ensure Prophet results files are secured for audit regulations the standard results are encrypted and cannot be read as input files for DCS. They will appear to contain random characters when trying to read the results as text files (i.e., .fac, .proj, etc)

 

To read tables and results outside of the Prophet interface go to Prophet > Run Settings > Results > Read Results File and tick the appropriate box to create plain text results files. Re-run the job and Prophet will create additional text files that can be read into DCS and other applications without encryption.

Using SQL Server connections via ODBC All

DCS allows you to use SQL Server as the source for Input Tables and as the destination for Output Tables. To enable SQL Server for either purpose you will need to setup an ODBC connection to an existing SQL Server database. 

1) Open the Windows Control Panel, go into Administrative Tools and then Data Sources (ODBC)

2) Add a new User or System DSN and select the SQL Server driver.

3) Give the DSN a name and enter the SQL Server Instance details, then select the default SQL Database where the tables are located.

 

Once the ODBC connection has been created you need to select the correct Run Settings in DCS.

 

If you want to use SQL to provide Input Tables:

1) Open the DCS project and navigate to the Run Settings page

2) Under the Run Settings > Input panel, change the File Type to DATABASE.

3) Change the Data Source to the named DSN connection created above.

4) Save the changes to the Run Settings

5) Create a new Input Format. Use the Browse button when prompted to enter a name and this will list all of the tables in the database. Select the table you wish to read from.

Note: If the browse button doesn’t appear to do anything then cancel the Input Format, re-select the Data Source and save the Run Settings again.

 

If you want to use SQL as the target for your Output Tables:

1) Open the DCS project and navigate to the Run Settings page

2) Under the Run Settings > Output panel, select the ODBC Output Data Source.

3) Save the changes to the Run Settings.

4) Open the Output Format and for the File Format option set ODBC to Yes.

5) Save the Output Format.

 

If you have any problems with the ODBC data source then try linking an Excel document to the ODBC connection. If this works you can then direct DCS to read from the Excel file and this will “chain” via the ODBC connection into the SQL Server.

Unable to load DCSDATES 8.2.3

Issue:

When upgrading to DCS 8.2.3 from a previous version and attempting to run a program, the following error occurrs during the DCS run: “Unable to load DCSDATES80.DLL”.

 

Solution:

Use the EXPLICIT_DECLARATIONS option at the beginning of the code to help identify any variables that need to be declared in advance of being used within the main code block. Once the variables have properly declared with their data types it should resolve the DLL error message.

Copying highlighted code deletes highlighted code 8.2.3

Issue:

In DCS, in the code, they highlighted a selected part of code in order to copy it and clicked on Copy on the home ribbon. The highlighted text then disappears out of the code. The copy button seems to act as a "cut" button. The program code is then marked as changed (the star appears in the tab), when they haven´t actually changed anything.

 

Solution:

This was identified as a bug and fixed in DCS 9.0.0.

 

Workaround:

As a workaround, use CTRL+C command to copy your intended code.

Troubleshooting DCS code with the MESSAGE command 9.0.*

Introduction

A common cause of errors in DCS runs is an unexpected change in the format or contents of the Input Files that do not meet the syntax required by the input format. For small input files it can be relatively easy to identify a syntax error in the data, but for larger files where there could be millions of rows and hundreds of columns it is very difficult to identify where the rogue data has appeared.

 

Identifying the row of data causing the error.

In most cases DCS will report the row number within the Input file that has caused a syntax error. However, if your data does not contain line numbers or is simply too large to open with common text editors then it can be very difficult to identify the offending row.

 

An easy solution is to use the Message command to write the current record number and an identifier field into the runlog just before the using the OUTPUT command. You can then review the runlog to see the last record that was successfully processed, and the next record should be the one that caused the error.

 

Identifying unexpected values.

Reporting the row number within the Input file is only helpful if the data type of that row is incorrect. In some cases the value will be of the correct data type and will not cause an error, but the value itself may be incorrect within the expectations of the model or code. For example, where a mortality rate is expected to be a fraction in the style 0.0123 to indicate a 1.23% chance of mortality, but the data in the input file has been incorrectly entered as 1.23 and accidentally indicates a 123% chance of mortality. Error checking can be included into the code to identify such unusual values. For example:

 

IF Mort_Rate >1 THEN MESSAGE("Excessive Mortality Rate detected in row: " + STRVAL(RECORD_COUNT))

 

Identifying incorrect selections.

Another common mistake is when a limited choice is required but the data-entry clerk mis-types the entry. For example, where Policy_Risk is a choice of “Safe”, “Low” or “High” but the clerk has accidentally entered “Hjgh” (note the J). You can easily include a check in your code to highlight any incorrect selections without heavily modifying the code. For example, assuming the code appears as…

 

IF Pol_Risk = "Safe" THEN…
ELSEIF Pol_Risk = "Low" THEN…
ELSEIF Pol_Risk = "High" THEN…
ENDIF

 

You could quickly add an additional line before the END IF to report any values that do not meet the required selection.

ELSE MESSAGE ("Pol_Risk does not match Safe, Low or High options in record: " + STRVAL(RECORD_COUNT))

Return to Top

Excel 

Title Versions known to be affected Article
ResultsReader DLL could not be loaded
8.x

After upgrading Prophet and Excel Reporting tool, when Excel is opened, you may get an error “ResultsReader DLL could not be loaded”.

 

The issue could be due to the required components to get Prophet Excel Reporting running not being installed or one of them is missing, typically the Microsoft RunTime DLLs.

This is most likely if Excel Reporting has been installed separately from the full Prophet installation.

 

Please check if these items have been installed. They will be necessary for Prophet Excel Reporting to work properly:

• MS .NET Framework V4 Microsoft

• C++ 2010 Runtime DLLs (32-bit)

• Microsoft C++ 2010 Runtime DLLs (64-bit) [only for a 64-bit machine]

ResultsReader DLL could not be loaded 8.x Before uninstalling, please open Excel and go to Excel Options | Advanced and scroll down to the General section and delete the path contained in the 'At startup, open all files in:' field. Once that has been done, uninstall the Excel Reporting via the Control Panel and reinstall.
Performance issue on Citrix All

The refresh of an Excel spreadsheet is taking considerably longer under a Citrix hosted environment when compared to running off a local machine. For example, a sheet that takes a few seconds to refresh locally can take up to an hour over Citrix.

 

To resolve this issue:

• Open Excel (or any of the Office programs) and select File > Options.

• In the Options dialog box, click Advanced.

• In the list of available options, select Disable hardware graphics acceleration.

• Click OK.

• Restart Excel

Failed to Create Query Sheet 9.0.1 only

Opening Prophet Queries on 64-bit Excel will cause an error reporting a missing file.

   "Failed To Create Query Sheet: Could not lead file or assembly"

The file can be located in the 32-bit folder:

   C:\Program Files\SunGard\Prophet Excel Reporting\x86\Sungard.WPF.Themes.dll

To resolve the error, copy the file from the x86 folder into the x64 folder.

Query Group is missing from the Excel ribbon Any

Ensure none of the add-ins have been disabled. Open Excel options, click on “add-ins” and in the drop down menu at the bottom of that screen, select “disabled items” and click on “go”. Enable any of the following add-ins:

- ExcelRPM

- Perms

- ProphetQueries

Return to Top

Glean 

Title Versions known to be affected Article
Mortality Table Requirements All

Glean tables such as the Expected Mortality rates need to be in a basic table format.

 

They could (but do not need to) be in the specific "Generic Table Format" used by Prophet, such as below:

Lapse Rates for ... projection – created

!2 PROD_NAME 0 1 2 3 etc.

* C_WEND 0 10 8.5 5

* C_IA_ 0 0 0 0

* U_MIP 0 12.5 10 8.5

* U_IB_ 0 10 10 10

 

Alternatively, the table could be in a “Column Format” such as:

Term Assurance Premium Rates - Table Created

* 20 10 0 1.52

* 20 10 1 1.04

* 21 10 0 1.53

* 21 10 1 1.05

 

Where the “middle” columns represent age, term, sex (smoking status etc..) and the last column represents the rate to be used.

 

You can also have a simple “base” table which just contains the “base rates” and use separate tables for the multiplicative factors for each type of risk. These can then be combined within Glean by utilizing the custom field functionality.

 

An example of using both Prophet Generic Tables, and standard mortality tables can be found in the Glean Life Example.

Compiling fails with no error messages Any

This may happen to our Glean users who do not have Prophet installed on their machines.

 

To rectify the issue, undertake these steps:

1. Apply write-access to folder and all subfolders: C:\ProgramData\SunGard\Prophet80\DCS

2. Put 7za.exe in any folder that is in their system path. (7za.exe can be found here - <C:\Program Files\SunGard\Prophet Worker\bin32>.)

Unexpected error: Retrieving the COM class factory 8.2.3

If the client encounters a “An unexpected error has occurred.. Retrieving the COM class factory for component CLSID…” error when trying to run Glean, copy the files listed to this path on their machine - <C:\Program Files (x86)\Common Files\SunGard Shared>.

 

-ProcLog32.dll

-ProphRes80.dll

-RegisterWin32.bat

-RegisterWin64.bat

-TableController80.dll

-xerces-c_3_1.dll

 

The batch files need to be run as administrator (if this fails, run from a command prompt in the location they copied them to.) Please contact the Prophet Helpdesk to obtain a copy of the missing files.

Return to Top

Licensing 

Title Versions known to be affected Article
Service hasplms failed to start
All

When attempting to install or update the Sentinel HASP runtime you encounter the error message “Service 'hasplms' failed to start. Verify that you have sufficient privileges to start system services”.

 

Either the user does not have Local Admin rights on the machine or the Sentinel runtime is not installed properly. To install the Prophet application on your machine, you should have Local Admin rights on the machine (network or domain admin rights are not sufficient).

 

If you already have Admin rights then it is the Sentinel Runtime component which is not installed properly. Uninstall Sentinel Runtime from Control Panel >> Programs and Features and re-install it from the Prophet DVD/Setup folder >> Tools >> HASP >> HASPUSerSetup.exe.

 

You can also download the latest “Sentinel HASP LDK” Runtime files from SafeNet:

http://sentinelcustomer.safenet-inc.com/sentineldownloads/

V2C Error: HASP LM is not running
All

When installing the HASP Run Time the following error displays:

“V2C Error: HASP LM is not running ErrorCode: 34 4 4232 33”

 

This can be caused by a restriction on user PC that is stopping the service from being started automatically.

 

1) Check that user has full administrator access for the account that is installing the software. Open the Control Panel > Administrative Tools > Computer Management.

2) Go into the menu for System Tools > Local Users and Groups > Group > Administrators. Open the administrators group and check that user account is listed here.

3) IT will need to confirm that they do not have any Group Policy restrictions that prevent services from being started automatically.

4) Check that the service for Sentinel HASP License Manager is set to Automatic and is currently started. Under the Recovery tab, change First/Second Failure options to Restart the Service and under Dependencies it should list the TCP/IP Protocol Driver.

Vendor extension not found All

One of the possible solution is you will need to check by going into the Control Panel > Admin Tools > Services and ensure the “Sentinel Local Licence Manager” service is in started mode and make sure that its kept on automatic mode in the startup type. Once this is done Please reboot your machine and check.

Transferring a licence Up to 9.0.3

The Prophet licence security system, Sentinel HASP, locks the activated licence onto the physical host machine where the activation key is used.

It is not possible for clients to move the licence to a new machine without the assistance of the Prophet Helpdesk. Attempting to move the licence by copying or transferring the hard drive or disk image to a new device will cause the licence to become disabled.

 

Transferring a licence to a new machine

The process of transferring a licence is the same for both Machine-locked and Floating-network licences.

In order to transfer a licence you need to create a C2V file from the old machine and send this to the FIS Prophet Helpdesk.

 

Gathering a C2V file (client to vendor)

Click on START > All Programs > SunGard iWorks > Prophet Professional > Configuration > HASP Remote Update System Utility.

On the tab for Collect Key Status Information, press the button to Collect Information to generate the C2V file.

The helpdesk will then send back a V2C file to remove the licence from the old PC.

 

Applying a V2C file (vendor to client)

Click on START > All Programs > SunGard iWorks > Prophet Professional > Configuration > HASP Remote Update System Utility.

Go into the tab for "Apply License Update", locate the V2C and click Apply.

You will then be prompted to create a C2V receipt file, which you need to return to FIS Prophet Helpdesk to confirm the licence has been removed.

 

Note: If you do not have this installed on your Licence Server you can run the application directly from the Prophet installation DVD, located in the directory…

PP/PE 902 ...\ProphetLicenceManager\SunGard\Prophet Licence Manager\Hasp

PP earlier ...\Prophet\SunGard\Prophet Professional\SungardRUS.exe

PE earlier ...\EPWeb\SunGard\Prophet Enterprise\SungardRUS.exe

Glean ...\DCS.Net\SunGard\Data Conversion System\SungardRUS.exe

 

Once the FIS Prophet Helpdesk have confirmed a valid licence has been removed we will then supply a replacement licence activation key that can be used to activate the new machine.

Return to Top

Prophet Enterprise 

Title Versions known to be affected Article
HTTP Error 404.7 – Not Found
All

If you use a customized IIS installation on the web server you may find that PE causes errors when trying to upload a workspace using BITS. The PE Studio log files will show “HTTP status 404: The requested URL does not exists on the server.” If you attempt to browse to the PE BITS application subdirectory URL via your web browser it will report “HTTP Error 404.7 – Not Found. The request filtering module is configured to deny the file extension.”

To resolve this you need to add the default file extensions and verbs to the IIS settings. Browse to the IIS site \ Virtual directory properties and open the Request Filtering options. The default settings should be:

Ensure “Allow unlisted file name extension” and “Allow unlisted verbs” are turned on.

However if this is not possible, a more granular solution is below:

 

Under the File Name Extensions tab there needs to be an entry for files that have no extension. This can be done by adding an entry using the single dot “.” with no extension text.

 

For the HTTP verbs in the Filtering page, add the following verbs and set them to True in the Allowed column. Note these are case-sensitive:

GET

POST

CONNECT

DELETE

TRACE

HEAD

PUT

OPTIONS

 

Once the IIS settings have been saved, browse to the PE URL location again and the error should now change to “HTTP Error 403.14 – Forbidden”, which is the correct default message as directory listing is forbidden. You should now be able to upload a workspace via BITS without errors on the web server.

Entity is located in a different compartment
All

When a PE User tries to Cancel / Delete Jobs, the user may get an 'Entity is located in a different compartment' error message. This happens due to the job is processed in one compartment and user is trying to cancel that job for which user do not have access.

 

Resolution:

1. Login with the user who has access to the compartment in which the job is started.

2. Cancel or delete the job.

Error enumerating BITS jobs
All

The error is seen in PE studio immediately after the initial logon.

 

This has nothing to do with the PE BITS option, but is due to the clients "Background Intelligent Transfer Service" BITS being disabled on the machine running PE Studio.

 

This service is set to automatic (delayed) by default on Windows 8, and presumably other previous windows versions, and is generally used for downloading updates, but may be disabled in custom client builds.

 

The resolution is to start the service, and close and restart Studio. The error message should now not appear.

How to Export a PE Workspace for investigation by the Prophet Helpdesk
All

There are several changes that need to be performed to prepare a workspace for exporting to an external system for investigations and trouble-shooting. The first step is to make a duplicate of the current workspace so that changes are only applied to a copy and not to the live/production workspace.

 

Export the workspace as normal and then Import this as a new workspace back into the PE environment. If you have a test environment then consider uploading the new workspace here and not into the production system. Give the new workspace an appropriate name that clearly separates this from any production workspaces.

If any mistakes are made while changing the “export copy” the workspace can quickly and safely be deleted and a new version imported without having to worry about affecting live data.

 

Working with the “export copy”, delete any Job Templates, Jobs, Run Structures and Run Settings that are not required to reproduce the issue being investigated.

Open the Run Settings and navigate to the Tables tab. Crop any links to network paths and mapped drives from the table location directories. Remove any unused paths and change any required paths so that they point to the local workspace directory.

For example, if the Model Point Files are located at \\FileServer01\ProphetData\MPF\*.pro then change this to only \MPF\*.pro.

Create a new folder on your desktop and copy the required tables using EXACTLY the folder names and directory structure as shown in the Table paths. Also include any customized Mortality Tables that have been added to your worker machine’s standard tables.

Navigate through the remaining Job Templates, Jobs, Run Structure and Run Settings to remove any Parameter values and replace these with fixed values used during the submission of the job that requires investigation. For example, if your Run Settings are set to use a parameter for the simulation numbers and your job was submitted with the parameter “1,10,100,1000” then you need to change the selection to use the fixed simulation values instead of reading from the parameters.

 

Export the “export copy” workspace and send this along with the tables folder to SunGard, along with instructions on which Structure and Run Settings pairing are needed, or which Job/Template are used to re-create the calculation run that is being investigated.

You can import this “final export” workspace, including all of the supplied tables in the workspace folder, to confirm with your own test run to ensure that you have supplied all of the required files and settings needed to correctly reproduce your issues.

Gathering Worker Log Files All

Enabling the PE Log recording

 

Introduction

To help debug infrastructure errors within the PE environment you may be asked by the Prophet Helpdesk to enable the PE Worker Logging on your machines. These changes need to be made to every machine used in the calculation run, including the Task Master and all Worker machines.

 

Viewing the contents of the ProgramData folder

By default the Windows operating system will hide the C:\ProgramData folder.

To be able to view this, open Windows Explorer and go to Tools > Folder options > View.

Enable the option for “Show hidden files, folders and drives”, which is located under the group “hidden files and folders”.

Disable the option for “Hide extensions for known file types”

 

Editing the log configuration files

Browse to the C:\ProgramData\SunGard\Prophet folder. This is where the configuration files and the log reports are stored.

 

Make a backup copy of the Worker folder. The configuration changes we are about to make will affect the performance of all calculation runs and will significantly increase the amount of disk space used. It is very important that the log configurations are removed once the required files have been gathered.

 

Once the backup copy has been made, edit the following files in the Worker folder using a text editor such as notepad. Note that not all files will be present under some installations.

AdminServer.config.xml

Broker.config.xml

Fastrun.config.xml

TaskMaster.config.xml

TaskMasterHPCv4.config.xml

Worker.config.xml

Within each file, search for the following line: <log_level>ERR+WARN</log-level>

Edit this line to change from Errors and Warnings to “ALL”: <log_level>ALL</log-level>

 

Also edit the Runtime.config.xml file. Look for the line below and set this value to 9.

<CreateCoreDumps>0</CreateCoreDumps>

 

Due to Windows security restrictions you may be prevented from saving the changes directly to the ProgramData folder. In this event to can save to the desktop and then manually copy the new file into the ProgramData folder, overwriting the old files.

 

Creating the log files

Restart the worker machines for the configuration changes to take effect.

 

Start your Prophet run and allow this to proceed as normal. The extra information logging will increase the overall runtime of the job and increased the amount of disk activity seen on the worker machines. Ensure that you have enabled the XPS runlog option and supply this file to the Helpdesk along with the worker logs.

 

Once the job is completed browse to the C:\ProgramData\SunGard\Prophet folder and zip the entire contents. Name your zip files to help identify which was the assigned Task Master machine and which are Worker machines. This process needs to be repeated for each of the machines used in the PE job.

 

Once you have gathered the runlog.xps and all of the worker logs, send them together to the Prophet Helpdesk for analysis, along with screenshots of any error messages that have appeared on any of the machines.

 

Restore the config files

Once the logs have been gathered and supplied to SunGard you need to restore the original config files to the workers. This will prevent further logging from consuming excessive disk space on the workers and will improve the runtime performance back to the original settings.

Configuring E-mail Notification All

To configure Email notification in Prophet Enterprise, first login to Prophet Enterprise Web server.

• Open Prophet Grid Administrator

• Go to File Management --- Scheduler and click on server name.

• on right hand side, click on "Schedulerservice.exe.config" and make below changes

   <add key="alerts" value="true"/>

   <add key="alert-mail-interval" value="0.0001"/>

   <add key="alert-smtp-server" value="SMTP server FQDN"/>

   <add key="alert-mail-sender-domain" value="Sender domain"/>

   <add key="alert-mailing-list" value=""/>

• Click on "monitormanger.exe.config" and make below changes

   <add key="alert-smtp-server" value="SMTP server FQDN"/>

   <!-- Domain name for the Enterprise.Prophet sender of alert emails -->

   <add key="alert-mail-sender-domain" value="Sender domain"/>

Limiting cores on a Worker All

The maximum usable cores can be set for workers from Prophet Grid administrator in TaskMaster.Config.xml.

1. Login to PEG Admin.

2. Navigate to File Manager Tab. Then Click on the Worker machine under ‘Workers’ in left hand pane.

3. In right Hand Pane Click ‘System Files’ tab click on ‘CECommon/TaskMaster.Config.xml.

4. In Right and below pane you can mention max number of workers that can be utilized in between ‘<max-local-workers>0</max-local-workers>’.

For Example if you want to limit cores up to 30 then the field will look like this:

‘<max-local-workers>30</max-local-workers>’.

5. Click on Upload Changes and select the workers to which you need to upload the made changes.

Job locked by user All

If the job monitor service encounters a problem such as a disconnected Task Master during a run it is possible for the job to become “stuck”. Users will encounter the following message when trying to clear the job from the queue:

PE Job Locked Error: Job- “job name” Locked by: <user> - Business Administrator{userid} Locked on

 

As a first step, please log out of PE Studio and try again when you log back in.

If that does not work, please check in the SQL database in your domain database and "jobs" table.

See if this job still appears I the table. If so, you can open this table in edit mode and delete the row with this job. Refresh the PE Studio screen and the stuck job should disappear.

Scheduler Service keeps stopping All

This may be happening due to EPCLIENT account is not able to communicate with the Configured Databases.

 

Solution:

Navigate to C:\ProgramData\SunGard\Prophet80\Prophet.Logs on the WEB Server.

1.Open the Monitor Manager Logs file and check for the error. In this case you will receive below error:

“Login failed for user 'EPClient'. Reason: The password of the account has expired.”

2.Login to the Detabase server and check the Status of EPClient account. Please unlock it if its locked.

Password Recovery for the Service Account All

It is possible for the IIS website administrator to recover the PE service account password, using a feature of the IIS services.

 

First open a command prompt and change to the inetsrv directory.

   Cd C:\Windows\System32\inetsrv

Run the following command, where PE9APP is the name of your app-pool, and this should return with the password value of you service account.

   appcmd list apppool PE9APP /text:ProcessModel.Password

 

If you do not know the name of the app-pool you can use the following command to get a list of the pools available

   appcmd list apppool

 

Note this password recovery option is a feature of IIS and may be removed in future patches and updates to the Microsoft operating system.

Compiler for a new PE Domain All

To set up the compiler in PE Domain please follow below steps:

1. Login to the PE Studio with newly created PE Domain URL.

2. Navigate to the Administration--> Domain Setting.

3. Go to Compilers. In Right Hand Upper pane you will see list of supported compilers.

4. Move the Installed Compiler from Supported compilers to Installed compiler list

5. Click Ok

 

Once this procedure is completed then only users will see the list of Compilers in Job Templates to select.

SuperAdmin account disbabled All

You will need to edit the 'dbo.User' table from PE Domain Database > Tables. 

Look for the specific user account in name column, and see for the value in Disabled column. If it

is True, change it to False.

Once the above change is done, the client can login using their existing password.

Unable to create Templates or Jobs All

Error: Loading New Job Template failed: Object reference not set to an instance of an object. 

Cause: This error is caused due to wrong configuration of Parameters in PE Studio, for example a user logging out while creating a new Parameter.

Solution: Perform below steps to resolve this issue:

1. Login to PE Studio and Navigate to Administration--> Parameters

2. Check for recently created parameters. If these recently created parameters are locked or not correctly Configured, please unlock or correct these parameters and test creation of template or Job.

3. If still issue persists, please delete these newly created parameters and then check functionality.

Uploading a Workspace with Nested Structures 9.0.0 + To allow the upload of workspaces containing Nested Structures, the following changes will need to be made:
1. Open Prophet Studio and connect to the PE domain
2. Click on the Administration tab and open the Domain Settings screen
3. Navigate to the Upload panel and change the options for Structure Links and Inner Run Settings from “None” to either “Selected” or “All”.

This will allow the users to include these elements when they upload their workspace to PE.

Worker Status "Not Responding" in PEGAdmin All

Error :

MonitorManager Log in Application server:

WARN Failed to ping Broker on <Server Name>:9008. <Server Name>:9008 failed to respond

 

Solution:

If a machine is set to “not responding”, it usually means the monitor manager on the application server cannot contact the broker service on that worker machine, consequently it won’t be able to update the health and status (i.e. alive or busy) of that particular machine on the database.

If you think the machine is no longer part of a run and it should be available, we would recommend checking for taskmaster and worker processes that have not been terminated on that worker machine, if you find this is the case it could be a problem with the model you are running, network related, virtual worker, etc.

 

To resolve this issues follow below steps :

1. Restart the Prophet worker services on the worker macine and check for the same.

2. Remove the worker machine from PE Grid Admin and try to re-add them from the list of available machines. And see if they are now available.

3. Re-starting the broker (worker machine), will terminate any worker related processes and help to clear the cache / temp memory and refresh the system resources that are running on that machine and make the machine available on the grid assuming there are no network or firewall issues.

If still the same issue persists then we would recommend you to uninstall and reinstall of PE worker application on the Worker server.

 

Note: Once you uninstall the PE worker from worker machine first delete Prophet Worker folder from Program Files\SunGard and then re-install the PE worker.

Error getting File Security when uninstalling a worker 8.2

Issue:

While uninstalling the PE Worker below error is occurred:

Error getting File Security: C:\ProgramData\Sungard\Prophet\GetLastError: 5

 

Cause:

Folder and its child folder at path C:\ProgramData\Sungard\ permissions are incorrectly set. Account who is performing uninstallation do not have permissions to access C:\ProgramData\Sungard\ path.

 

Solution:

1. Manually apply the permissions to the C:\ProgramData\SunGard\ Folder and forcefully apply to the Child Folders as well.

2. Refresh and check permissions, if permissions are changing please verify Group Policy. Check the CREATOR OWNER permissions are pushed though GPO, if yes, please remove that and perform GPUPDATE.

3. Once these permissions are applied correctly uninstall the PE worker.

This will uninstall PE Worker successfully.

SQL 2012 'Always On' feature All

If using the new Always On feature of SQL 2012, a few additional steps need to be carried out in order for the SQL databases to be usable when they failover.

 

If these haven't been done, PE will appear to work, but when submitting a job it won't do anything. There will be no error message.

 

1. The DB username must be created manually on the secondary node with the same SID.

2. Permissions must be assigned on the databases when they are failed over for the first time. The account needs to be mapped to each respective database and be part of the DBO schema. The user account needs datareader and datawriter rights as well as Connect and Execute permissions on the databases

3. The 3 custom error messages need to be injected into the master DB. Further details are available from the helpdesk.

Return to Top

Prophet Professional 

Title Versions known to be affected Article
Reducing the variables calculated
All

By default all variables for all products are calculated during a run. However it is possible to choose a smaller subset by altering the ‘Calculated Variables’ in the Run Structure. The specified variables are calculated as well as any variables that they depend on. These dependencies will be calculated across products.

 

It is possible to choose a variable group, a specific variable or all cumulative variables.

 

When specifying a subset such as variable group X93 – Stochastic Output below, Prophet will only calculate the variables that are present in the X93 – Stochastic Output variable. Note that variables can exist in multiple variable groups.

 

To see which variables are included in this variable group go to workspace objects, variable groups and navigate to the specific variable group.

Reducing the variables in the results files All

If you want to restrict the variables that are included in the results files then you should use the run setting as well or instead.

 

You can again restrict these output variables to a specific variable group or to cumulative variables only. By using this setting you will not reduce run time but you will reduce the file size of the results set.

 

Variables that are not included in the ‘calculated variables’ (including dependent variables) as set in the structure will not be able to be reported on. Similarly variables that are not included in the ‘Variables in Results Files’ cannot be reported on either.

Cannot see results in Diagram View All

Possible Causes include:

1. You are using the incorrect version of Diagram View. If you cannot see your results you may be using Product Diagram View instead of Results Diagram View.

2. You have opened Results Diagram View but have not made any selection. A variable, a run number, a sub-product and a results type must be selected before results can be viewed.

3. The results viewing pane at the bottom is obscured by other viewing panes. This is particularly common for laptop users.

4. The variable’s results were not set to be calculated in your run structure options or they were not included in the results’ display options in your run setting.

Compare report font size All

Some users may find that the font used in the Compare report is too small, too large, or does not remain the same between reports.

 

To resolve this, after you run the comparison report, go into the Compare menu, under the Settings group, and select the Options button. On the subsequent screen click the Font button. Here you can set the font size back to a more readable size.

 

If you find that Prophet repeatedly changes the font again, set the value as above and then ask your IT department to remove your permissions to edit the following Registry key to prevent further changes:

   HKEY_CURRENT_USER\Software\SunGard\Prophet\Beyond Compare\Properties

 

This will lock the current font value in place and force Prophet to keep that setting, but this solution will prevent the user from being able to change the font again until the registry permissions are restored.

Comparing objects between versions All

You cannot compare libraries or workspace contents from different versions. For example if you want to compare the products / libraries between a Prophet 8.0.2 workspace and Prophet 8.2.3 workspace. The workspaces in which the item you are comparing must be of the same version.

 

To perform the comparison, you can make a copy of the 8.0.2 workspace, upgrade that to 8.2.3 and then run the compare against the same version.

 

However, be aware that READ_RESULTS will be automatically changes to READ_RESULTS_LEGACY when upgrading a workspace between certain versions of Prophet.

 

Any changes made to variable definitions as part of an upgrade process are logged in a text file stored in the same location as the Prophet workspace. Further details of those can be found in the release notes accompanying each release.

Blank compiler list All

After installing “Visual Studio Express for desktop applications”, when Prophet is launched the Compiler List is blank and the following errors are displayed if you try to start a run:

“Default compiler selected is not valid for this architecture”

“Failed to load implementation of ERBuild.dll.”

 

The compiler must be installed before the Prophet Worker component. Use the Windows control panel to remove the Prophet Worker component and then re-install this. The existing compilers are detected during the installation process and should then appear on the list in the Prophet options.

 

All installations should be done using administrator rights.

Adjusting CPU / Memory usage All

The Prophet worker process attempts to make full use of any CPU and Memory resources available. The needs are highly impacted by the type of model and “split-by” choices. For example a stochastic model with a single product will make heavy use of a single core if split-by-product, but this would allow the core to use all of the available memory. If you split the run by simulation then all of the CPU cores will be used but the total memory will be divided between the cores, resulting in less memory available for each set of calculations.

 

If your calculation job requires more memory than is available from your machine, or if you have set the Run Settings to use minimal memory, the working data will be pushed to disk to attempt to reduce the overall memory being used. Disk-based calculations are significantly slower than in-memory calculations.

 

In-memory dynamic (newer ALM) models usually do not split well over multiple cores. ALS models typically require large amounts of memory compared to the cores available.

 

To ensure best performance we strongly recommend the use of dedicated worker machines for Prophet to maximize the use of memory and CPU resources and avoid any potential performance loss caused by other applications running on the calculation machines. To avoid having runs with data cached to disk you can change the scope of the job to use fewer cores, calculate fewer products, or to report a reduced set of variables.

Using fewer cores will allow the disk to run more smoothly and increase the memory available per core, hence this may speed up the response from cores that remain in use. You should not allocate more cores than the physical machine has to offer, and for larger servers we recommend to reduce the overall available cores slightly, for example in a 48 core server only allocate 46 cores to Prophet and keep 2 cores free for Windows to perform operating system tasks and to allow administrators to login and monitor the server’s performance.

 

Standard monitoring tools such as Windows Task Manager will allow you to view the Disk, CPU and Memory activity to help decide on the resources needed to run the job at its most efficient state.

Debugging a SameAs Product All

You can set and manage breakpoints to debug 'Same As' products as well as for ‘Master’ products. This can be done either outside of a debug run or while a debug run is being processed.

 

First identify the Master product associated with the ‘Same As’ product you wish to debug. Open the Master product, identify the required variable and set the breakpoint in that variable. Then in the Breakpoints window, right-click on the breakpoint and use ‘Change breakpoint product’ to change it to the same-as product. You will only be offered a list of products sharing the same Master product.

 

It should show up as enabled. At this point if you close the master product (and all other Prophet windows) and then click on the Same-As product and select ‘run’ while in debug mode it would stop at the breakpoint.

Conditional Debugging All

The Debugger in Prophet allows you to step through calculations line by line, or variable by variable, to follow the flow of your model. Breakpoints can be introduced to allow you to start the step-through in the middle of the calculations. These can be based on first reaching certain lines in the code, or when certain conditions are attained, such as a variable having a specific value.

There are two issues around setting those break conditions:

 

Option of beginning the Debugger when a variable is equal to a specified value:

When you start a debug run you have the option of beginning the debug process when a variable is equal to a specified value. Although you can select text variables from the list, the debugger will ignore text conditions and will begin the debug process from the first record.

If you would like to condition the debugger on a text value we suggest you use an integer flag within your code and select this flag to begin the debugger.

 

Option of beginning the Debugger based on a variable:

If a variable is of data type Number, the number in the Equals box must contain a decimal point, even if the number of decimal places is zero. If a variable is of data type Text, the value must be entered without any quotation marks. The value is case-sensitive.

Debugging with Print_To_File All

The PRINT_TO_FILE command can be used within Prophet extended formula code to report on values being passed into and out of calculations, READ_GENERIC_TABLE and CALL_DLL functions. This can assist with debugging issues where an error is being caused by a specific row in the model point or input data, or where unexpected values are being returned in the results.

 

For example, when using the READ_GENERIC_TABLE command you can use the following code to record the value being searched for and the value being found which you can then manually compare with your table to identify problems with the source data.

   PRINT_TO_FILE("C:\Example.log","Y", “Searching for: ”, MyName, MyID)

   LUValue := READ_GENERIC_TABLE("GENTABLE", "N", MyName, MyID, "LUValue")

   PRINT_TO_FILE("C:\Example.log","Y", “Found: “, LUValue)

Note that using the PRINT_TO_FILE command will slow down the calculation time of your model and can produce very large log files, particularly when used within loops.

 

You can use a constant value defined in your code to toggle the reporting of investigative messages. This allows you to quickly bypass the creation of the log files in production runs and re-enable them for debug runs without having to re-write or add comments to multiple lines of code or quantities of variable definitions.

 

Create a Constant variable set to either 0 for standard runs, or 1 for debug runs:

   Constant FLAG_DEBUG = 1

Use the variable in the code to decide if a log needs to be written:

   IF FLAG_DEBUG = 1 THEN PRINT_TO_FILE("C:\Example.log","Y", “Found: “, LUValue)

 

You can use escalating levels of reporting based upon the value of the flag:

   IF FLAG_DEBUG >= 1 THEN PRINT_TO_FILE("C:\Example.log","Y", “Search: ”, MyName, MyID)

   IF FLAG_DEBUG >= 2 THEN PRINT_TO_FILE("C:\Example.log","Y", “Found: “, LUValue)

 

Setting the flag to zero will then bypass all of the logging activity for all of your production runs. You can also use enumerations to provide meaningful names to each level of reporting used by your flag.

 

It is good practice to remove debug code from the model before taking this into the final production system.

Older Database Version All

User's may see the following message when opening a workspace:

Error: Older Database Version

 

Solution :- You get this error message when trying to upgrade to the latest version or taking a back up of the workspace and then running it. Copy the PRW file into a new, empty folder and try opening that in the new version. If that works then copy only the Tables and Model Point files over to the new folder. Don’t copy over any other files such as the Workspace Files.

 

If copying only the PRW file doesn’t work then you will need to start a new workspace on the old version and use the IMPORT feature to bring in the libraries, products, accumulations etc from the old workspace. Once that is done you can then try copying the new workspace over to the new version of Prophet.

Error opening Results Diagram View All

When trying to open a product in Results Diagram View, users may encounter the error message "Results Diagram is only available after the product has been run."

Solution: You should not rename the prw file. If you rename it back to PROPHET.prw, you will be able to view the product within the results diagram view.

Error: Structured Exception All Structured exception errors typically happen when a variable has the wrong definition type. For example, when you try to pass an array into a constant variable. You should check any recent changes you have made to the product variables to ensure they are using the correct types and compare this with any variables being read for the library to ensure they are compatible.
Access is Denied All

The following errors may appear while running a PP job:

PreparationService::buildProduct exception: boost::filesystem::remove: Access is denied

PrepareBuildTask failed, boost::filesystem::remove: Access is denied

 

You may not have permissions to access the folder specified. Please have your IT check your permissions to the folder, and have them enable Full Control permissions.

 

If you are using a backup /restore of another user or machine’s workspace, the dll files from that workspace may not be compatible with the current compiler on your device.

 

Create a copy of the workspace folder in case this needs to be restored later. Delete all files except for the workspace PRW file, and all folders except for the Tables and Model Point files. Everything else, such as the Structure_Files, workspace_files and Runs folder should be deleted. You should be able to open the PRW file and run the workspace as normal.

 

If another user can run the workspace or backup and restore the same workspace on the same machine, the problem may be with your Windows User Profile.

 

Please have your IT department backup/delete your Windows Profile on your machine. If your IT department are using roaming network profiles they will also need to remove the network version of your profile to stop the faulty profile being copied back onto your machine after it has been reset.

 

When you log back in to the machine, it should then create a new profile with the correct permissions for accessing the files that you create. Prophet users should have Full Control over the workspace folder and all contents and sub-folders.

 

If you still have problems your IT department will need to work through the differences between the two different user accounts or machines to identify which settings are preventing you from accessing the files.

Tried to write to a disconnected socket All

The following error messages can occur during a Prophet run using remote worker machines:

Request to <machine:port> failed. Start.Job failure. Failed to parse definition.

When starting Run pair: Failed to put a message. Tried to write to a disconnected socket.

 

These typically appear when a required network communication port has been restricted either by the local firewall or by a remote firewall located on a network switch or router device.

 

 

Please check with your IT department to make sure the following ports are open on all devices used in the Prophet calculations.

 

Incoming firewall ports:

• AdminServer.exe (incoming port 9002)

• FBServer.exe (incoming port 3058)

• ProfBroker.exe (incoming port 9008)

• TaskMaster.exe (incoming ports 9100, 9200, ...)

• Worker.exe (incoming ports 9101, 9102, ..., 9201, 9202, ..., ...)

 

File and Printer Sharing ports:

• File & Printer Sharing Port 139 TCP/IP

• File & Printer Sharing Port 445 TCP/IP

• File & Printer Sharing Port 137 UDP

• File & Printer Sharing Port 138 UDP

 

Prophet Licence Manager

• HTTPS Port 443

 

Note that there is no upper limit on the port numbers used by the Task Master and Worker processes. The theoretical port limit is 10150 based upon 10 task master processes using 50 cores in the final job, however if a specific port fails to respond the task master might go beyond this limit as it tries the next port number range available. Opening the range between 9100 and 10500 should cover the majority Prophet Professional job requirements.

Failed to parse run definition All

When starting Run Pair: Request to XXXXXXXX : 9008 failed.StartJob failure, Failed to parse run definition

 

Solution :- You will encounter this error message when trying to run a workspace from the shared drive. You will need to make sure that the workspace folder has full permission to the users to at least R/W and also the machine account name of the user has been added onto the workspace folder.

Other workaround is that you can copy the workspace locally and run it.

Failed to read runlog summary All

When performing a run you may encounter the error “Failed to read run log summary, error and warning counts may be incorrect”.

 

This seems to be an access or permissions issue. The RunlogSummary.xml file should be located in the results\RUN_* folder and the worker machines need to have Full Control over this file.

 

Close the Workspace and Prophet Professional if they are open on the machine. Use an Administrator account to give full control access to the User who is running the workspace on the following folders:

   C:\Program Data\SunGard\Prophet80

 

If you use a named service account for remote worker machines then you also need to ensure that service account has the same Full Control access to the folders.

 

If you use more than one compiler you should make sure you have full permissions on the following registry key:

   HKEY_LOCAL_MACHINE\SOFTWARE\SunGard\Prophet

 

Also, give full permission to the Workspace folder. It could be your Local drive or network drive, depending upon from where you are accessing the workspace. You may need help from your IT team to provide rights to the particular folders and registry settings, particularly if you use service accounts on remote worker machines.

Firebird Server is not running on the Workspace File Server All

It is possible to continue receiving the warning that Firebird Server is not running on the Workspace File Server after the Prophet Hosting and Firebird Server tools have been installed to the Workspace File Server.

 

This is caused by a setting in the Workspace folder that was recorded before the workspace hosting tools were installed to the file server.

 

Use Windows Explorer to browse to the Workspace folder and edit the .det file in a text editor such as Notepad. Search for the entry “AccessibleFBServer3058=No” and change this to “AccessibleFBServer3058=Yes”, saving the changes once this has been done.

 

This will allow the users to open the workspace from the Workspace Hosting server without further warnings.

General database error All

General database error: Exception class: EProphBOMException Exception address: 007F3DB7

Workspace may be corrupt.

 

The main reasons for the appearance of this error are terminating the Prophet Professional process without closing the workspace first, network issues while closing the workspace stored on a network location, someone making changes on the workspace stored on a network location while you are closing the workspace, any files deleted accidently from the workspace.

 

Try deleting the .lck file and .det file in the workspace folder and try opening the workspace again. However, if it still does not work, please try restoring the backup of the workspace from when it was last working.

 

Best practice for not corrupting workspaces hosted on a network is to use a Prophet workspace hosting installation. This involves installing workspace hosting and Firebird 2 server. The workspace hosting option is not installed from the wizard, but from the .msi found in the ProphetServer folder. The Firebird 2 server can be found in the tools folder on the installation disk.

Goal Seeking loops are limited to 50 iterations All

The number of Goal-Seeking loops are limited to 50 iterations by default. This can be adjusted by editing the configuration files.

 

To increase the limit, go to C:\ProgramData\SunGard\Prophet\Worker and open the file “Runtime.config.xml” using a text editor such as Notepad.

 

The settings can be found on the following lines:

<!-- Maximum number of goal seeking loops. Default is 50. -->

<MaximumNumberGoalSeekingLoops>50</MaximumNumberGoalSeekingLoops>

 

Note that most goal-seeking calculations quickly converge, and increasing this limit is unlikely to improve the convergence of your model.

Missing mortality tables All

We provide standard industry expected mortality and morbidity tables in Prophet-compatible format to customers as part of the standard Prophet Professional (and Glean) installations. The tables which are provided are the standard expected rate mortality and morbidity tables in international varieties, which are freely and publicly available.

For tables which are copyright or subscription protected, we are unable to publish these into a Prophet format and therefore the conversion of this data into Prophet readable format is the responsibility of the customers. Certain tables, such as those published by the Institute of Actuaries in the UK, are no longer available publicly and therefore we are unable to provide a Prophet-formatted version either with a system release or as a stand-alone option. Should this data at some stage become freely and publicly available, not requiring any subscription and without copyright protection from the author, then we will consider the possibility of adding to our current tables pack.

Adding mortality tables All

Prophet can read mortality tables either from the standard tables folder, or from the tables folders specified within your run settings. The standard tables location is specified in your Prophet options, in the Calculation Engine section. This is typically located in the Program Data folder. Adding additional mortality tables here will make them available to all structures, but the update does need to be done on all Prophet machines. Alternatively the tables can be added to the folder specified in the Run Setting of each workspace, making it easier to share between machines, but requiring all workspaces to be updated. All worker machines will also need to be able to access the mortality tables location.

 

Default location of the Mort folder:

The default installation location for Prophet Mortality tables is C:\ProgramData\SunGard\Prophet\Mort.

Note the ProgramData directory is a hidden folder and the user will need to reconfigure their Windows Explorer settings via Tools > Folder Options > View > Show hidden files, folders, and drives.

 

Changing the Mort folder location within the PP 9 interface:

To change the default location of the Mortality folder, open the Prophet Options and navigate to the Calculation Engine section of the Machine Options. Press the Edit button and enter or browse to the new path.

NO_CALC changes 8.2.2+

From version 8.2.2 of Prophet, a number of enhancements have been made to the behaviour and sequencing of variables which referenced variables containing NO_CALC. These variables will report an error message to show that the required value has not yet been calculated.

This may cause older models to fail after they have been upgraded to a later Prophet version. This is common in models that assumed a zero-value would be instead returned.

To resolve this you can use the Diagram Viewer and Runtime Diagnostic files to identify where variables are being calculated and take steps to correct any dependencies.

We recommend that in code that you refer to NO_CALC with one of the 3 parameters (or a combination). This will help to maintain code clarity. It will also help the Prophet system.

Attempt to read uncalculated value All

ERROR: PCR003 Attempt to read uncalculated value

 

This error appears when the variable currently being calculated is dependent on the result of another variable that has not yet been calculated.

 

You can use the Diagram Viewer and Runtime Diagnostic files to identify where variables are being calculated in the wrong order and take steps to correct the grouping to ensure dependent variables are calculated first within the same group as the variable that require them or in an earlier group.

 

You can use the GROUP_AFTER(variable) function within a variable’s formula to queue the calculation order.

 

It is also possible to insert interim calculations to force a dependency such as:

{normal code} + 0 * VAR_NAME

Where the VAR_NAME is the variable from a later group that needs to be calculated first. Note that multiplying the variable by zero will always result in zero, so no numerical change is being performed to the normal code results, but the future variable will then become a dependency to be calculated in this earlier calculation group.

 

Also note that the + 0*VAR_NAME and GROUP_AFTER can force more variables to change their calculation order than the one variable that you reference. It can also force variables into rebasing or calculation loops. It should therefore be used with care. These functions will also be ignored if the system is unable to apply these enforcements due to a contradictory need elsewhere in the code.

Removing the vertical line in Code Editor All

When you’re in the view ,click on ‘Editor Options’ > ‘Advanced Editor Options’ and look at the ‘Right Edge’ section. Set the value of "Edge Column" to 0.

 

Unwanted Projection results files 9.0.1

When running a Prophet job where the run setting options are set to not produce Projection files; if you run the job using the multiple or default machine settings (and you have configured the machines) then the projection files will still be created.

 

For Professional, running with the local machine option will correctly refrain from creating the projection files.

 

When working with multiple worker cores it is required that the task master compiles the .proj files in order to process any accumulations and generate any text results required; as such it is impossible to avoid creating these files when multiple cores are used as it would impact the calculation process. It is still possible to manually delete the files after the run is complete, or to use an end-of-run batch file to help automate the process.

Projection term is greater than maximum All After adjusting the Maximum Projection Term in the Run Setting for the run, set the product in the run to Regenerate. Adjusting only the Maximum Projection Term will not update previously compiled products. The product needs to be regenerated in the run as well.
Reusing Run Number deletes results All

By default Prophet will delete any existing accumulation results when you run a subsequent run pair on the same run number. It does this to ensure that you are not using results in the second run that you did not intend to.

In fact it only deletes accumulation results that exist that are at a level number at least as big as the lowest level number being run. For example if you have results for an accumulation, ‘A’, at level 20 and if the run ran products and accumulations from level 10 onwards then the results for A would be deleted. However if your run was running products and accumulation from level 25 then results for A would not be deleted.

It is possible to turn this setting off such that accumulation results are maintained. However if you do this then you will need to make sure that you are using the correct results and not some old results that you didn’t realise were still there. The setting that controls this is ‘Delete Accumulation Results from Lowest Level’ which can be found under the Advanced Results Settings on the Results tab within the run setting.

Run button disabled for Push-to-PE All

Problem: In Prophet Professional user could not do "Push to PE" runs, because the "Run" button was disabled.

 

Error message "PE Run in Progress. Cannot open Results file"

 

Solution: User must close the PE Job Monitor to exit PE session before they are able to login from another session to perform Push-to-PE runs. Clients’ Citrix administrator may need to identify any disconnected session to ensure the user has correctly close all previous instances of the PE Job Monitor. Please note that PE Job Monitor application may still be running on the task bar and user needs to right-click the icon to exit this properly.

Unable to see licensed libraries 8.2.3+

The client is unable to see licensed libraries and components after installing Prophet 8.2.3 or 9.0.

With this release, the installation of the standard workspaces was moved to a second source - the Solution Installer. Please check that you have also installed this component.

Error validating formula 8.2.x only

Error: Validating formula: System Error: Access Violation

 

This occurs when modifying the definition of variable that includes “ SAME_AS_PROD_NAME” [when used: READ_GENERIC_TABLE("CODE_" + SAME_AS_PROD_NAME]

 

Cause:

This has been confirmed as a bug and will be fixed in a future release (fixed in 8.2.3)

 

Workaround:

As a workaround, please change the relevant variable to be a core variable and make the changes to it in the library. Ensure that you regenerate the product when performing a run.

Missing DLL Entry Points All

Error :- Missing DLL entry points: Unable to find GetProdAnalyserFacade in ProdAnalser.

 

Causes :- This error normally pops up when you have not installed a Compiler on your machine or the existing compiler installation has been broken.

 

Solution :- Please install a supported compiler as per the Prophet version update or try to repair the existing one to resolve the issue.

Time Index Error All

This message will be displayed if there is any attempt to read a value for a time period before zero or after the end of the projection period.

 

It is normally caused by a formula which contains references to variables at incorrect time periods. For example, the following formula for SUM_ASSD_PP would result in this error at time t=0 because the code is attempting to read from t-1:

SUM_ASSD_PP(t-1) * 1.05^(1/12)

 

The formula needs to be amended along the following lines:

IF t=0 THEN

10000

ELSE

SUM_ASSD_PP(t-1) * 1.05^(1/12)

Return to Top