Removing duplicate files in GridFS

gridfs_remove_dups.js

// Run this script if the following command fails to clear duplicates
// > db.fs.files.ensureIndex( { md5: 1}, {unique: true, dropDups: true} )

// detect duplicates in db.fs.files
m = function () {
    emit(this.md5, 1);
}
 
r = function (k, vals) {
    return Array.sum(vals);
}
res = db.fs.files.mapReduce(m, r, { out : "myDupesCollection" });
 
// delete duplicate files AND their corresponding chunks
db.myDupesCollection.find({value: {$gt:1}}).forEach(
    function(obj) {
    var cur = db.fs.files.find({ md5: obj._id }, {_id: 1});
    var first = true;
    while (cur.hasNext()) {
        var doc = cur.next();
        if (first) {first = false; continue;}
        print ( doc._id )
        db.fs.files.remove({ _id: doc._id });
        db.fs.chunks.remove({ files_id: doc._id})
    }
})
 
// delete dups collection if you don't need it anymore
//db.myDupesCollection.drop()