Import operation

pakira

Hi All,

I am facing performance issue with DFC import operation. It is taking around 20 Sec to import a 300KB pdf document. Below is the code -

public String doDDFSImport(IDfSession session, String docPath, String destFolderPath, String objectType) throws Exception {
     DfLogger.debug(this, "Method Start doImport", null, null);
     String newObjectId = null;
     try {
    IDfClientX clientx = new DfClientX();
    DfLogger.debug(this, "destFolderPath:  "+destFolderPath, null, null);
    IDfImportOperation operation = clientx.getImportOperation();  
    operation.setSession(session);
    IDfFolder folder = null;
    folder = session.getFolderByPath(destFolderPath);

       if (folder == null) {
     DfLogger.debug(this, "Folder or cabinet " + destFolderPath+ " does not exist in the Docbase!", null, null);   
        throw new SystemException("Folder or cabinet " + destFolderPath+ " does not exist in the Docbase!");
    }
    operation.setDestinationFolderId(folder.getObjectId());
    // check if file or directory on file system exists
    IDfFile fileObj = clientx.getFile(docPath);
    if (fileObj.exists() == false) {   
     DfLogger.debug(this, "File or directory "+docPath+" does not exist in the file system!", null, null);
     throw new SystemException("File or directory "+docPath+" does not exist in the file system! for object type : "+objectType);
    }
    // add the file or directory to the operation
    IDfImportNode node = (IDfImportNode) operation.add(docPath);
    node.setDocbaseObjectType(objectType);
    node.setNewObjectName(fileObj.getName());
         
    boolean importResult = operation.execute();
    DfLogger.debug(this, "importResult:    "+importResult, null, null);
    String errorMessage = null;
    StringBuffer sb = new StringBuffer();
    sb.append("Operation failed: ");
    sb.append(errorMessage);
    sb.append("; Errors are: (");
    try {
     IDfList errors = operation.getErrors();
     for (int i = 0; i < errors.getCount(); i++) {
      if (i > 0) {
       sb.append(", ");
      }
      sb.append("[" + i + "] ");
      IDfOperationError error = (IDfOperationError) errors.get(i);
      sb.append(error.getMessage());
     }
    } catch (DfException ex) {
     sb.append("UNABLE TO GET ERROR MESSAGES ("+ ex.getStackTraceAsString() + ")");
    }
    sb.append(")");
    DfLogger.debug(this, "error:   "+sb.toString(), null, null);
  
    IDfList myNodes = operation.getNodes();
    int iCount = myNodes.getCount();  
    DfLogger.debug(this, "Number of nodes after operation: " + iCount, null, null);
    for (int i = 0; i < iCount; ++i) {
     IDfImportNode aNode = (IDfImportNode) myNodes.get(i);
     DfLogger.debug(this, "Object ID " + i + ": "+ aNode.getNewObjectId().toString() + " " + iCount, null, null);
     newObjectId = aNode.getNewObjectId().toString();
     DfLogger.debug(this, "Object name:    "+aNode.getNewObjectName(), null, null);
    }
   }catch(Exception de) {
    DfLogger.error(this, "Error during import", null, de);
    throw new Exception(de);
   }
  
   DfLogger.debug(this, "Method End doImport", null, null);
   return newObjectId;

  }

Can you please suggest how can I improve this performence. Thnaks for your help!

Kind regards,

Sumanta Pakira

 

DCTM_Guru

There are many places where performance can be affected.  Some of this may be affected by your network itself.  Have you tried pinging from your computer to the content server to see how many hops it takes?   Another place to check is the actual database itself.  Do you have auditing enabled?  If so, then importing a document will cause an audit trail to get generated.  I would check with DBA and see if they can run any stats against the database.

Assuming its not the network or database related, have enabled DFC tracing?  This will enable to place timestamp for each DFC call.  This can also help identify which method is taking the longest.