Extract Command Failing in One Site but not Another

Usage and Config Platform
1

I have a fixlet that copies the Windows 11 source to a local directory. I use bfarchive to create a compressed version of the Windows 11 source. Then the EXTRACT command extracts the source when the fixlet runs. It has been working without an issue for several months. Now I copied the fixlet to a new custom site and now it is failing on the extract. It seems to extract most of the files before it fails.

2

This issue was caused because the newly created site name was to long. The name of the site where the EXTRACT command is working is 8 characters long. The new site name is 18 characters. This length went over the limit. I am not sure exactly what the limit is. After creating the new site with 8 characters, the EXTRACT command was successful.

3

Windows or Linux or ??

4

I just tried on a custom site with a 16-character site name and had no issue. I’m not sure the length of the site name is the issue.

5

I had a space in the site name “XXXXX Workstations”. Could that be it?

6

Possibly? What exactly did your extract command look like? May need to quote the filename if it contained spaces.

I’ll try to reproduce but seeing a copy of your extract command might help.

7

A Custom Site with spaces in the site name, should represent those spaces with underscores on the client’s filesystem:
image

Below is a copy of the original and extract file as viewed on the client:

 Directory of C:\BES\Client\__BESData\CustomSite_This_Is_A_Test\__Download

06/23/2025  10:09 AM    <DIR>          .
06/23/2025  10:09 AM    <DIR>          ..
06/23/2025  09:29 AM             1,843 win_extract.bfa
05/19/2025  03:21 PM             5,656 win_extract.txt

This was generated by this actionscript:

begin prefetch block

add prefetch item name=win_extract.bfa size=1843 sha1=3f449bc8c3e88b1b05a586eba09bd95057cbd245 url=https://127.0.0.1:52311/Uploads/Custom/win_extract.bfa sha256=76044904c78127c8432bf70c1e130c69f6e2c0edf2396f11f7566f17b5a8615d

end prefetch block

extract win_extract.bfa

Normally we don’t need to reference the site name, directory name, or download folder with the ‘extract’ command - it just assumes sitename\Download - but I think we can also specify paths if we want to, which is why I’d want to see what your extract command looks like.

1 Like
8

‘extracting most of the files before it fails’ is kind of triggering as well.

Windows itself has a character length limit on paths by default; I think most of my test systems have that length removed, something that can be done with a registry edit that Python installers tend to do by default.

I’ve seen cases where a few Microsoft installers only work when executed from a root directly. I think it might have been Visual Studio or something like that, which had a really deep directory structure that didn’t work when you stick Program Files (x86)\BigFix Enterprise\BES Client\__BESData\CustomSite_Sitename\Download in front of it because it was too long.

I wonder whether your specific package has such a deep directory structure on it that is right at a limit where your 8-char site name is ok but your longer site name combined with the archived directory paths become just too long?

9

It looks like the path name length is 256 by default.

It looks like the longest path with this package is 200 characters. So, I don’t know how it got over 256 characters, but it must have.

C:\Program Files (x86)\BigFix Enterprise\BES Client__BESData\CustomSite_xxxxx Workstations__Download\Windows 11\sources\dlmanifests\microsoft-windows-ie-clientnetworkprotocolimplementation-migration