UG NX 4 Cad File Network Storage

[kdd282]

I store my CAD files on a network server. I'm looking for a way to speed up the search of an initial load because of the system has to search my defined network drives in the load options section. We have a lot of files and it can take a while to load.

Does anyone have successful method in tweaking the performance of UG when searching the directories to load all the necessary part files?

I already use the gigabit network cards and devices.

Thanks for your help.

 

[ewh]

File segregation according to project is how I do it. If you have a project with a huge number of files, you may need to segregate further, by area or subassemblies. This will of course require updating of the load options to reflect which files need to be opened. Also, you could change the scope of your load options, using partial loading or structure only, then open only the files you need.

"Good to know you got shoes to wear when you find the floor." - [small]Robert Hunter[/small]

 

[kdd282]

ewh,

Thanks for the post. We use these parts in other projects and don't want start keeping muliple copies of the files. We are taking a steps towards a PLM system but not quite there yet.

I was think of some type of a 3rd party program that would help the search function.

Thanks.

 

[looslib]

Separate those files that are standard parts out to a separate folder structure and load them first, then load the rest of the files from a separate folder. You can separate files by product line and use different load option files for each product.

Do you use part versioning? That has an impact on file loading, too. Also, NX will always load the first occurance of a file it finds, so multiple locations for the same file does not work.


"Wildfires are dangerous, hard to control, and economically catastrophic."

Ben Loosli

 

[hudson888]

What you need is something that works robustly with a search algorithm. Your file naming and versioning system comes into play here as well, especially if you're using versioning rules. The first thing to avoid is having doubled up file names in different places even if you think they're copies of the same thing. The second is the order in which you search directories and indeed the number of directories and sub directories that you're searching through. Avoid where possible a long chain of directories and subdirectories that have to be searched through, it both takes too long and risks things being easily lost, misplaced or worst of all doubled up.

If you have different naming from one project to the next or for differently used data sets then that is ideal because you can afford to store the files separately with less risk of doubled up file names.

With directory structures you probably want to reflect your organisational workflow so that released files are kept separately from work in progress. Using windows system investigate setting owner file permissions so that files that a team are working on can be stored in a collective directory without risk of them overwriting one another. Then you want to be able to search that directory usually first and the release directories later.

Change the names of files as they go up a release as soon as you begin working on them. In that regard look into versioning rules and if you have the choice use the military system. You need to be able by name to regulate and control your assemblies content otherwise you'll have a persistent state of chaos.

Best Regards

Hudson

www.jamb.com.au

Nil Desperandum illegitimi non carborundum

 

[hudson888]

P.S.

If you do use versioning rules then if there are two versions of the same core file name stored in a top directory and a subdirectory then unless you specifically search the sub directory first the system will always find whatever version is stored in the top directory. Having an directory structure inappropriate to your work flow can see the system regularly load older versions rather than newer ones.

Best Regards

Hudson

www.jamb.com.au

Nil Desperandum illegitimi non carborundum