Locked *.tmp files when polling images with Rulerunner (aka Quattro) and VScan task in Datacap 8.1
Hallo, greetings from The Netherlands,
I use a Rulerunner service for automatic actions (VScan, Profile, Export). I have some problems with 'timing' in VScan.
When I copy the files into the input folder (image), the service is picking up the files. But sometimes it's stopping and it is leaving a *.tmp file. Looks like some timing issue with writing of filesystem, and Datacap reading to quick. Causing a 'Abort', and stops Rulerunner service. (I have configured 'Stop after error')
Anybody experiencing same problems? And maybe some tips?
Thanks, Sander.
This isn't a problem with 8.1. It has been a problem with prior releases of Datacap. I found that the out of the box vScan actions are not production ready due to the issue you're experiencing. I've only used the out of the box vScan actions for demos. You will need to customize a vScan action to:
- look for a trigger file that is written to the ingestion path once the image has successfully complete prior to ingesting the image or
- create a custom vScan that evaluates a files created date. If the date is older then a configurable property (i.e., 15 min), then you can assume the file is complete so you could ingest those images.
I hope this helps.
Tony
Hi S.Baars
This is a common issue. There is a post regarding it in Datacap 8.0 which was supposedly fixed. However it is still prevalent.
My recommendation would be to use the C# VScan library (Available as VScanSample from developerworks site). With some modification this action library performs better than the VBScript default one.
I believe it to be some form of contention issue (file access denied, corrupt etc)
I have not seen any instances of .tmp files with a C# VScan... comments from others would be appreciated.
Regards
Tom.
Thanks, no easy fix, but fixable...
Thanks for your time and sharing your experience.
Thanks Tom,
I will look which one is used (We use FastDoc/FastApp as template). And than I will do some extra testing.
Thanks for your time and sharing your experience.
Hi S.Baars
We used to have the same issue in our implementation. We have ftp jobs from our branches pooling the multi image tifs from the branch servers to the central server where datacap process it with vscan. The Vscan was configured to get only one tif for each batch.
When the FTP was slow, it created a lot of tmp files ( that where the processed tif incomplete, rename it and open it to check).
We fixed this issue changing the image folder to another one and having a new job moving the files.
This job moves the tif files to the new images folder but with a changed extension that datacap ignores. After it renames it. Because this last action its really fast we issues with the tmp files disappears.
(Note: Whe didn't change the ftp jobs to upload the files with a temp extension because we needed to change more than 600 jobs manually)
One more issue when you got the tmp files issue is that some times with the correct timing Datacap process an incomplete tif file (Last image of the tif truncated).
Regards,
Juan.
Juan, thank you for the tip. Always nice to learn different solutions from other people. Will discuss it with client to see if this strategy is the way to go. Thanks, Sander.