Deploying a constantly changing file

(imported topic written by SystemAdmin)

Hi All,

We have a CSV file that is extracted from TAMIT that we are using to brand our servers with various attributes. This file is updated daily and I have a fixlet that is set to run on a daily basis to reset these values. The fixlet is working fine, but what I am seeing is that the updated file is not getting set to the target. I have the fixlet deleting the file from the __Download folder before it is downloaded, but I think what is happening is that since they are cached on the relay, the updated file is not getting used.

Another thing to note is that the file is on an internal web server (AIX) and I have to use it from there as there is no method to move it to the TEM server uploads folder.

Anyone have any ideas on how to get around this without having to start a new action every day?

In TCM land, I would have used what was called the Data Moving Service :slight_smile:

Thanks

Martin Carnegie

Gulf Breeze Software Partners

http://www.gulfsoft.com

(imported comment written by SystemAdmin)

There are basically two approaches, each with their pros and cons.

  1. Action regenerator: This is probably the more officially sanctioned approach. It involves scheduling the action regenerator untility on an interval to check a folder and create a signed manifest of ‘good’ files that are valid to download. In this approach, you are trusting whatever is put into that folder and added to the manifest. The agent does check the signature before using the download, so corrupt files or man-in-the-middle issues are avoided. However it isn’t quite as flexible as the ‘old school’ option.

  2. ‘Download now as’ approach: This is fast, flexible, and avoids manifest creation hassles (but lacks the signature validation). In your actionscript do the following syntax:

download now as somefile.ext http://someserver.company.com/uri/somefile.ext

continue if {exists file “__download\somefile.ext” whose (size of it > 0)}

Note that this goes directly to the target web server and bypasses your relays (which can be desireable under certain circumstances).

There are certain select circumstances where it is extremely useful to be able to do relevance substitution in the download URL. You can do that with a normal ‘prefetch’ style download, but you can with this approach. For example:

download now as <file_name> http://someserver.company.com/uri/{variable for location, dept, etc.}/somefile.ext

continue if {exists file “__download\somefile.ext” whose (size of it > 0)}

In certain special circumstances, relevance substitution in the URL has been extremely helpful.

(imported comment written by SystemAdmin)

Hi Jon,

I will check out both those options. I am thinking that since it is such a small file and deployed to mostly local servers, that option 2 might work better for me. Still, I will look at #1 just to make sure :slight_smile:

Thanks

Martin Carnegie

Gulf Breeze Software Partners

http://www.gulfsoft.com