duplicity-team team mailing list archive

[Henrik Farre <question291337@xxxxxxxxxxxxxxxxxxxxx>]

Question #291337 on Duplicity changed:
https://answers.launchpad.net/duplicity/+question/291337

Henrik Farre posted a new comment:
So... I don't know what to do next :)

I have tested different chunk sizes using this example code:
https://www.dropboxforum.com/hc/en-us/community/posts/205544836-python-
upload-big-file-example using a very unscientific method where I just
added datetime.datetime.now() before and after the upload, and then
converted the numbers to MB/s:

1Gb in 4Mb chunks: 10min, 44 sec ~ 1.58MB/s
1Gb in 16Mb chunks: 7min, 53 sec ~ 2.15MB/s
1Gb in 150Mb chunks: 5min, 45sec ~ 2.97MB/s
1Gb in 250Mb chunks: 5min, 23sec ~ 3,17MB/s
1Gb in 300Mb chunks: 5min, 15sec ~ 3,25MB/s
1Gb in 350Mb chunks: Connection reset by peer
1Gb in 500Mb chunks: Connection reset by peer

150Mb is max recommended by http://dropbox-sdk-
python.readthedocs.org/en/master/moduledoc.html?highlight=files_upload_session_start#dropbox.dropbox.Dropbox.files_upload_session_start

The test script has ~15secs between uploads, e.g. there is no network
traffic in that time.

When I run duplicity there is no network traffic in ~30secs, and if
apply this patch: https://github.com/dropbox/dropbox-sdk-python/pull/54
I get a timeout:

dpbx code error "HTTPSConnectionPool(host='content.dropboxapi.com', port=443): Read timed out. (read timeout=30.0)"
Attempt 1 failed. ReadTimeout: HTTPSConnectionPool(host='content.dropboxapi.com', port=443): Read timed out. (read timeout=30.0)

So I think I just have to suck it up, change chunk size to 150 and let
the backup run until it is done :)

-- 
You received this question notification because your team duplicity-team
is an answer contact for Duplicity.