Large Document Handling

, ,
21

Manohar
Thanks a lot for the feedback.
You are very helpful.
Sathish.

22

Need help with Large Doc handling in 6.01

I have done the following:

  1. Configured WmTn with tn.BigDocThreshold and tn.tspace.location
  2. Configured WmEDI with EDIBighDocThreshold and tn.tspace.location
  3. I can see my document in my tspace
  4. When I review the contents fo the document in TN console my content
    name=EDIdata,
    Storage Type=tspace
    Storage Reference = com.wm.app.tn.tspace.Reservation_FileImpL

In Developer when I view the bizdoc I have the following
bizdoc/Contentparts[0]/PartName=EDIdata
bizdoc/Contentparts[0]/StorageType=tspace
bizdoc/Contentparts[0]/LargePart?/MBoolean=true
bizdoc/LargeDocument?/MBoolean=true

According to the documentation from this I can conclude that I have configured Large Doc handling correctly. But when the getContentPartData is executed I get following Exception.

“Error retrieving content of content part EDIdata of document vg00… The exection message is: Content for Part EDIdata of document vg00… does not exist in TSpace.”

Looking at the message I notices that the case on tspace has changed to TSpace. When investigating the WmTN properties.cnf file the entry
tn.tspace.location=C:\Data\LargeDocTemp has been changed to
tn.tspace.location=C:\Data\LargeDocTemp once the server has been restarted.

Has anyone seen this before? Why is my tspace parameter being rewritten incorrectly and why is the exception referencing tspace in a different case.

TIA

23

Chris,

Maybe your error message says TSpace but in the config file it should be ‘tspace’ only.

Please make sure your config file has this format. Note the double ''.
tn.tspace.location=C:\Data\LargeDocTemp

24

Manohar,

It is tspace in the config file and tn.tspace.location is set to
C:\\Data\\LargeDocTemp but once the server is started it changes to C\:\Data\LargeDocTemp

25

FYI

Problem solved.
Apparently placing a Save/Restore pipeline before the getContentPartData removes the reference from StorageRef in the bizdoc because it’s a reference to the object.
Hmmm!

26

try passing the bizDoc.Content to pub.io.bytesToStream
Then map the stream output to the input data.stream of wm.EDIINT.send

That might do it.