Renames causing data storage issues

[SK_11616132]

For many years, uses have been using rename to "save-as" parts in Creo managed by Windchill with no regard to the parts being used in other assemblies and released data. Windchill fortunately manages this and fixes the pointer when the assembly is opened.

We are trying to archive our data and get out of Windchill to use Creo native. These renames are causing HUGE issues because the exported assembly from our Vault calls for the old name but we are exporting the renamed part and it doesn't know where to point. For 1 to 1 instances, we are able to manually fix the list, but when renames have happened multiple times or when the old name is reused again later the data is too jumbled for logic.

Has anyone delt with this or have any advice? Is this something re-vaulting can help with? Our company has not done that at any frequency over many years.

[TomU]

You **CANNOT** just grab data from the vault and expect it to be correct.  Windchill manages all kinds of things at the database level (parameter values, file names, etc.) that are not reflected in the physical files in the vault.  The only correct way to pull CAD data out of Windchill is through the workgroup manager - the workspace in the embedded browser inside Creo.

[avillanueva]

I was going to echo what @TomU stated. Think about what Windchill does during a rename. Its quick and records the rename event in the database. What it does not do is go to the Creo files in the vault, open and update references. This ONLY occurs when files are brought out to a workspace inside of Creo application. This means that if I just download a file that had its reference renamed, and I do this multiple times, its gets corrected each time. Same is true if I have something open in session and a rename occurs on the Windchill side, I get altered and it fixes is real time. 

In your case, the only supported way to get the CAD data out is to bring it out via a workspace and export from the workspace. What kind of volumes are we talking about? Workspaces can handle thousands of objects so it might take a few passes but you get get it all. 

[SK_11616132]

Thanks for the reply @avillanueva. Definitely aware of all the file handling Windchill accomplishes on the back end. Unfortunately our latest released and in work data is around 1 million files... so our scale has definitely made this harder.

[d_graham]

@TomU is correct. You have to somehow run the cad data through WGM.

I've worked with a few companies recently that were moving away from Windchill and wanted to get all their CAD Data out of Windchill. These companies had hundreds of GB of data even if we were getting only the latest rev.iteration of each CAD object.

I wrote a utility that took advantage of Packages to get the data out of Windchill.

Here's how we did it.

Neither had a CAD worker setup, which is mandatory as the CAD worker will process the cad files thru the Work Group Manager when the Package is zip.

So, we setup CAD workers and I wrote a utility that automatically to do the export.

Wrote and ran an SQL query to get the total file size we planned to export.

Made sure the server had plenty space for these files.

I created only one Package manually.  This package will be reused over and over again by a loop in the utility.

The utility does the following:

1. Unlocks the Package, if it is locked.
2. Added CAD objects to the Package.
3. Created a delivery.
4. Zipped the delivery.
5. Find the Zip file in the Windchill file vault.
6. Download the Zip file to another Windows folder.
7. Delete the delivery.
8. Loop back to step 1 until all data is extracted.

Once we had all the zip files downloaded, we ran another utility to extract the cad data from the zip file.

The company wanted the Windows folder structure where the cad data was to be stored to match the Windchill folder structure, So, we automatically took care of that too.

If you more info or help reach out to me, but rest assured this is quite doable.

David

windchill.developer@yahoo.com 

[SK_11616132]

This is very interesting @d_graham. Any approximation for how long that would take for around 1 million files? How big can each package be and does it gather related files? We used SQL to get our list and then just run a batch code on the server and that takes about 24 hrs... but now we are dealing with the rename issues.

[PG_9878904]

David Graham worked with me on a project very similar to this just 2 months ago. we had about 260k files, and 1600 job folders. and it took about 5-6 days overall. i ran into issues with queries not allowing multiple cadworkers at the same time, as well as search queries would get filed up and the tool would stop running and i would have to restart it. there were ways around it but i just buckled down and watched it over the course of a week instead of wasting time trying to make it more efficient. 

so a couple of questions:

do you guys have a cadworker or packages setup?

what version of whindchill are you on?

have you tried using the export tool and if so do you have the same result?

have you tried loading the problem jobs in a workspace, and does the problem still follow if you do?

his custom made script was able to run all the folders files through a cadworker and export them as a package (which essentially functions the same as putting a job through a workspace)

David Graham gave us a very fair quote and worked tirelessly to solve the issue with me, so if you are looking for outside assistance for a reasonable price i might suggest you go that route. 

[d_graham]

@SK_11616132 

Your questions have been answered by @PG_9878904 

I should mention I have a utility that will get files directly from the file vault. That utility took under 3 hours to get 400 GB of data. As has been mentioned, no guarantee on data quality if pulling straight from the vault.

The thing to compare is the date of the vault file's creation relative to when it was re-CADNamed (is that a word).

This includes, of course, all objects that reference the renamed file, not just the file that was renamed.

If the date of the file is before the rename action, May Day! May Day! 😀

The cad file needs to be processed by the WGM.

I have written a utility that finds the number of all CAD docs where rename was done AFTER the latest storage of the cad file too.

Having done this, I'd play it safe and use Packages.  That technique works.  It takes longer but so what.  At the end of the day, you'll have rock solid data.

[anursingh]

Hi @SK_11616132,

I wanted to see if you got the help you needed.

If so, please mark the appropriate reply as the Accepted Solution. It will help other members who may have the same question.
Please note that industry experts also review the replies and may eventually accept one of them as solution on your behalf.
Of course, if you have more to share on your issue, please pursue the conversation. 

Thanks,
Anurag