What functionality should I put in Amazon Glacier Uploader or how do you store your photo archive?

Question

Amazon Web Services

After the advent of Amazon Glacier, I decided to make a utility to work with it. I must say right away that I am doing it primarily for myself, and therefore, it will take into account, first of all, my use-cases. But perhaps it can be useful to someone else. In this connection, I have a question:

Who keeps the archive of photos?

I have a case like this:
1) All photo sources are stored in one folder. In this folder, folders are created with a name containing, as a prefix, the date of shooting, and then the name: “2012-05-30_Kebabs in the country”. I upload photos, immediately delete the defective ones, and in the future the contents of these folders are unchanged. Processed/sorted photos are stored in another folder. there is much less volume.

I can assume that someone is using a different case:
2) the photo folder contains a directory, such as: "Nature", in it, for example, "Forest", but there is already a folder with the date (invariant in content).

Therefore, I want to implement the following functionality:
for sources (according to case 1):
a) scan the folder with the source photos and compare them with the list of folders already archived on Glacier.
b) all new folders are archived and sent to Glacier. One folder - one archive file.
c) the list (source folder name, archive id on Glacier, date) is stored on S3.

for processed photos:
a) make a daily differential or additive archive with a basic (full archive) 1 time in M ​​days.
b) we upload the archive files to Glacier (or you can start on S3), and delete them from there after 3 * M days. or through 2*M.

any other suggestions for functionality?

Answer 1

I myself keep photos of the first case. Making a backup on an external hard drive.
I used to encounter such a problem: the
folder is called simply by date, because when I threw it off, there was no time to name it. I saved it, and then, on the computer, I gave the folder a name, and during backup, I already have to manually rename the folder on the external one, so as not to confuse the backup program. I think that it is worth adding a check for equality of folder names with the source photos and archived ones, and in case of renaming the name of the original, rename the archive (the date in the folder name remains unchanged).

Answer 2

For me, the first case with tags (forest, booze, flowers) in the file properties is more convenient. I look through Picasa, there is still such a thing as a virtual "Album" - links to specific pictures from different physical folders.
The only negative is that the picasa makes a turn only inside itself and for itself, if you view it with Windows, you have to turn it over. I am now slowly writing a utility that will resize photos up to 640x480 / 800x600 for uploading to the network, and probably immediately flip them so as not to do double work on flipping also in pica.
sir06will - in your case, it may be more convenient to somehow calculate the checksums of the file, and then it doesn’t matter how it is essentially called - the hash matches - it’s hard to get confused.