CollectionFS with GridFS data duplication?

[lpgeiger]

Can anyone with GridFS experience tell me if data duplication is to be expected?

I’m having massive duplication issues. 250MB becomes 1,500MB in the DB

I’ve started with 250MB of CSV files, that I need to put in storage for Meteor to retrieve at a future date. After uploading my total database size is 1,500MB (before, my it was just 20MB).

The relevant collections:
cfs._tempstore.chunks - 40 KB
cfs.files.filerecord - 168 KB
cfs_gridfs._tempstore.chunks - 473 MB
cfs_gridfs._tempstore.files - 40 KB
cfs_gridfs.filesStore.chunks - 537 MB
cfs_gridfs.filesStore.files - 40 KB

1 - Where is the remaining +/- 500MB going?
2 - Why is tempstore not clearing the duplicate chunks?
3 - Why is cfs_gridfs.filesStore.chunks, double the size of the original files?
4 - Is this GridFS behavior or just a CollectionFS pattern?

[robfallows]

GridFS uses MongoDB and MongoDB …

Mongo collection size, why so big?

I’m having a strange issue here where I have a collection with 2,200 documents in it that’s over 400mb. Further, I’m trying to render the whole collection client-side, which is obviously problematic. The thing I can’t understand is why it’s so large, as it’s storing very little text. I know why it used to be large, which is I was storing entire web page content (as stringified HTML) inside each document. When I did that, my collection totalled 600mb, which I thought was a lot, but there was a l…

So it could just be that. If you have replica sets on different servers and GridFS duplicates chunks for redundancy, that could also explain it.

(I know almost nothing about GridFS, so this is speculation on my part)