Ask the Expert: Best Practices – 7.x Installation and Upgrade

Did you know...When applying a new CAF/LIC file to a system that is already licensed and activated please follow these steps to ensure success:

When upgrading to a new version of Captiva consider upgrading to the latest scanner and ISIS drivers for the scanners in your field.  A great resource to get started is www.scannerdrivers.com.  You will be less likely to run into issues with your scanners because EMC mostly QA's new Captiva builds using the latest scanner drivers.

Did you know that page count sharing has been enhanced in Captiva Capture 7.5 and now applies to both the server licenses and Advanced Recognition licenses. Classification and Extraction now share page count between different servers in a ScaleServer group.

Upgrade to Captiva Capture 7.5 and take advantage of page sharing with Advanced Recognition in a ScaleServer group!

Did you know...   

Captiva 7.x still ships with a host of sample processes!  All the samples are documented in the help file:

Captiva Designer > Scenarios and Samples > Using Samples

Captiva Designer is installed with a collection of sample XPP files, profiles, recognition projects, document types, queries, and other configuration files. Sample files are installed on your desktop in the following directory by default:

:\Program Files\InputAccel\Client\src\Sample Capture System

To make the samples available for use in Captiva Designer, you need to create a capture system with the Copy Samples option enabled as described in Adding a New Capture System. All samples are copied to the folder created specifically for that capture system:

:\Users\ \Documents\Captiva 7.0\ \GlobalData\

Note: The subfolders included in the GlobalData folder are described in Understanding Capture System.

The copied samples are meant for demonstration purposes and are available for editing in Captiva Designer's embedded editors. For example, you can open a sample XPP file in CaptureFlow Designer and use it as a model for your own CaptureFlows. The example of creating a new CaptureFlow from a sample is described in Creating a Process Based on the BasicRescan Sample Process .

Sample files are installed with a collection of readme TXT files. The Readme First.txt file contains an introduction to samples and an overview of folders created for each capture system. Other TXT files contain descriptions of the provided sample processes located in the \GlobalData\XPP folder. The readme files are located in the following directory:

:\Users\ \Documents\Captiva 7.0\ \Readme\

The provided samples are as follows:

• AdvancedMultiDocCapture Process
• AutoIndexing Process
• BasicMultiDocCapture Process
• BasicRescan Process
• ExpenseCollate Process
• ExpenseCollateAdvanced Process
• FWTax1040 Process
• WSInputRescan Process
• WSOutputScan Process

Note: The AdvancedMultiDocCapture Process builds off of the instructions in the TXT file for BasicMultiDoc Capture.  Therefore it is recommended to become familiar with the contents of the BasicMultiDoc process instructions first.

can we customize the Captiva Web client?  is there any documentation related to that?  We are looking for integrating xCP with CWC.  Any thoughts on this?

Just to add some additional color to what Shaishiv has posted regarding setting up a cluster…

In my experience the easiest way to activate the cluster servers for an active/passive setup is as follows:

1. Install Cluster_CAF file on primary server. Obtain profile ID 1 from Admin. Apply LIC file on primary.

2. Force failover to secondary server and obtain profile ID2.

3. Submit both profile IDs to EMC Licensing to obtain both Activation codes.

4. Apply Activation code that maps to Primary on Primary.

5. Force failover to secondary and apply the other code there.

Hello Hari,

Thank you for your question. The Captiva Web Client cannot be embedded into an existing web application in the same way eScan’s ActiveX control was able to for example.  This is a HTML 5 based web application.  While unsupported, one could edit the CSS style - e.g. from webserver, locate the CSS class name and make changes like blue to red however chances are that any patches applied in the future will overwrite your modifications. 

As Captiva Capture Web Client requires an InputAccel Server, you can export data into a Documentum back end using the standard export module.

I suggest looking at the Captiva Capture help guide for version 7.5 for more information . You can find this via the “support by product” page on the support website:

https://support.emc.com/products/30391_Captiva-Capture

Did you know…

While Captiva 6.5 required scripting to do field validations, the best approach today is to leverage Captiva 7’s rule-centric
validation approach. 

Captiva 7’s Desktop and Extraction use a rule-centric approach to field validation which is quite different from the field-centric approach that IndexPlus, Validation, and earlier modules have used. The intent of this change was to reduce both the amount of scripting that is necessary to implement custom validation and the possibility for logic problems in tracking
field errors. To implement efficient validation logic, it is essential to understand the design and behavior of the rule-centric approach.  We highly recommend only using script where the rule-centric validations is not ideal to provide the desired function, such as modifying the content of a field.  For more information on this, please see our white paper located here:

https://support.emc.com/docu55856_Understanding-Validation-Rules-in-Captiva-Desktop-and-Extraction.pdf?language=en_US 

Please note that when InputAccel Servers are configured as a ScaleServer group, client modules must connect to one of the InputAccel Servers in the ScaleServer group by using the machine name of the machine hosting the InputAccel Server. If an IP address or the name “localhost” is used in the Server name field of the connection string, the connection to the server will fail.

If you use Captiva Capture for mission-critical applications, you may want to make sure that your Captiva Capture system is always available.

Here are some High Availability Best Practices to consider:

• At a minimum, connect the InputAccel Server machine and the InputAccel Database machine to

an uninterruptible power supply.

• Configure the SQL Server for high availability by setting up database mirroring and/or clustering.

Refer to Microsoft recommendations for advice and instructions.

• Configure the InputAccel Servers for high availability by using ScaleServer groups and

configuring them in an Active/Passive or Active/Active Microsoft Failover Clustering cluster.

• Run unattended client modules as services and configure those services for high availability by

enabling automatic restart on failure.

Did you know...  

If you are a user of the Western OCR Engine (which is a fantastic engine BTW) which is part of Advanced Recognition and you are upgrading to Captiva 7.1 or 7.5 from a prior release, please note that the default settings for all of the out of box Western engines HAVE CHANGED.

Your project will automatically pick up the new default settings of this engine when you upgrade your project.  Therefore it is recommended to evaluate the OCR output results you were getting with the previous release with the results you are now seeing with your upgraded project.

If you edit the out of box WesternOCR_Extended.reco for example (which will force you to create a copy of the engine) you will notice that the setting for TRADEOFF now has a default set to Fastest.  In prior releases this setting was defaulted to MOST ACCURATE. 

We have seen cases where results are not as accurate when the engine is set to fastest and by simply creating a custom instance of Western OCR with the Tradeoff set to ACCURATE you should see the same or better OCR results you were getting previously.  In my testing, MOST ACCURATE is still extremely fast and should be the preferred config.

When installing the IAS folder to a UNC path, if you get an error stating that the user account rights for the server has not been set and it may prevent the server from running correctly.

In this scenario, ensure the following:

     1. Run the server using a domain account.

     2. Grant the UNC directory for IAS folder to have full Windows permissions for that domain account.

     3. Ensure that every Captiva user that accesses and creates processes and batches are granted the Windows permission (use Windows tools) to access the IAS folder.

Please note that although supported, installing the IAS folder to a UNC path is not recommended

Hello,

We have only 2 days left for "Ask The Expert" event. I encourage everyone to actively participate and take advantage of this event.

Please feel free to reach out to us should you have any questions around "Best Practices - 7.x Installation and Upgrade".

Shaishiv

Don't forget to deactivate antivirus software on the InputAccel Server data directory. Running antivirus software on the InputAccel Server data directory and its subfolders will drastically degrade InputAccel Server performance due to the large number of files being written to the directory structure. In addition, some antivirus software intercept network traffic and can

interfere with InputAccel Server operation. In all cases, you should exclude the following directories and their subdirectories from antivirus scanning:

InputAccel Server data directory (by default, C:\IAS)

InputAccel Server installation folder (by default, C:\Program Files\InputAccel

\Server)

C:\ProgramData\EMC\InputAccel

Windows Temp folder (%TEMP%)

C:\Users\ \AppData\Local\Temp (where is the name

of a user)

Antivirus software is not designed to check in real-time the kind of volume and file size needed for a InputAccel Server to maintain full production throughput. This high volume of work tends to manifest antivirus software issues (usually hanging) that can in turn cause a production Captiva system to crash. The files in the directories for Captiva use are transitory;

that is, they exist only as long as the batch is in Captiva.