This is not a description of the implementation, it is a question for the community on what is better to implement.
There is a task:
Downloading to the server and storing digital images with sizes up to 1.5 GB, along with standard photos, as well as video and audio materials (hereinafter referred to as storage objects). It is necessary to organize more advanced storage than just storage in the file structure. Namely, generate thumbnails from images, including, at some moments, generate images for DeepZoom, do swf conversion for previewing video. Well, there are no questions with audio, you can listen to it and so on.
It is also necessary to provide a convenient interface for searching and working with storage objects, rather, it will be a web interface, and by the way, the first question is how much the web interface and, accordingly, the scripting implementation is applicable to solve such a project.
')
In fact, data loading, rather, the algorithm of operation will be as follows: source data in the form of images, audio and video are copied to a location on the storage server.
Further, a certain mechanism (script, executable) should essentially give the user a choice of which data to download from this folder, including with indication of certain parameters for a group of files (quality of digital images, what is obtained, what status they have). These parameters are not pulled out of the EXIF, because relate more to business logic than to technology, and therefore there must be an interface for setting these parameters.
Next, the user must select the folder where the objects are stored. This folder is selected either from the file structure or from some logical structure related somehow to the file structure.
Next, thumbnails are made from the selected graphic files, swf is converted from the video, the source files are moved to the file structure along the selected path, the thumbnails are also added to a specific address. In the database, records are created about the objects entered, their storage paths, and the parameters selected when loading objects are recorded.
Some additional data: the number of objects 2-3 million - this is still the foreseeable maximum that can be reached. More real numbers up to 1 million records. The storage of all this is also there - one of the options for HP EVA, although so far as a disassembled set of pieces of iron.
So the question is:
What OS in combination with other software is better suited to solve the problem?