Summary

This article describes how to determine the name and source (job, instance, and step) for captured file data stored on disk by adTempus. This procedure may be useful if, for example, you find an unusually large file and want to know where it came from.

More Information

Files captured by adTempus jobs (using File Capture actions on Responses, or using the Capture Console window mode for programs) are stored in the "data\CapturedFiles" directory under the adTempus program directory (e.g., "c:\Program Files\Arcana Development\adTempus\data\CapturedFiles").

All files are stored in a single directory and are named with a GUID (e.g., "{B32B74C3-AB67-4EF0-8E92-31E097B5D064}") rather than with their original file name and extension. The file contains only the data originally captured; it does not contain any information about the job that captured the file or about the file's original name.

Generally you will work with Captured Files through the adTempus Console. When you do so, adTempus displays the original file name and other information about the file. This information is stored in the adTempus database. When you view or save a Captured File, adTempus retrieves the data from the CapturedFile directory.

It may sometimes be necessary to identify a file in the CapturedFiles directory. For example, you may find an unusually large file in the directory and want to know which job produced it. There is no way to find this information using the adTempus Console, but the information can be obtained by querying the adTempus database:

  1. Run the adTempus Database Utility program (adtdbutil.exe) found in the adTempus program directory. adtdbutil will connect to the adTempus database and display an empty query window.
  2. Enter the following query: 
select 
f.originalfilename as 'File Name',
j.name as 'Job Name',
ehi.jobinstanceid as 'Instance',
case when ehs.oid=f.owner then ehs.stepnumber else null end as 'Step Number'
from job j
left outer join executionhistoryitem ehi on j.oid=ehi.joboid
left outer join executionhistorystep ehs on ehi.oid=ehs.owner
left outer join capturedfile f on (ehi.oid=f.owner or ehs.oid=f.owner )
where f.oid='{0A183A18-FD08-45DE-B8F6-FEC855575797}:fileguid'
  1. In the query, replace "fileguid" with the name of the file you are trying to identify, for example, "{B32B74C3-AB67-4EF0-8E92-31E097B5D064}" (do not include the double quotation marks).
  2. Press F5 to execute the query.

The tool will list the original name of the file, and the job, instance, and step that captured the file.

If no result is found this may be an "orphaned" file that has been removed from the job history but not purged from the disk. If this is the case, adTempus should remove the file the next time the adTempus service is restarted, or you may delete the file yourself.