How to Transfer Large Files to Atlassian

Atlassian's file service

To transfer a file via your browser:

1. In your browser, go to transfer.atlassian.com. Follow the prompts to log in using your Atlassian account. 
   
   
2. Your open support requests will be listed. Choose Add Attachments. 
   
   
   
   
3. Follow the prompts to upload your files.
   
   
   
   
4. Once the transfer completes successfully, check that the file size listed matches the size of the local copy of the file.
   
   
   
   
5. Repeat steps 2 & 3 for each file that needs to be uploaded.
6. Your support.atlassian.com issue will be updated each time a file has been uploaded. Once all your files have been uploaded, add a comment to your support ticket to let the support engineer know you're done uploading files. 

To transfer a file using via the command line:

1. After logging in to transfer.atlassian.com, choose Auth Tokens in the header to generate an authentication token. 
   
   

2. Use the example CURL command generated by filling the auth token you generated and your file access path to upload your file to this ticket as follows:

   curl -u NWMzNzcwY:xyz123 -X POST --header "Transfer-Encoding: chunked" \
       -F "files[]=@/var/atlassian/confluence/export/Confluence_support_2017-10-12-13-29-42.zip" \
       -F "files[]=@/var/atlassian/confluence/export/thread_dumps.zip" \
       https://transfer.atlassian.com/api/upload/PS-12345 -v --progress-bar | tee /dev/null

 transfer.atlassian.com is available for all of your open cases, so you can use it at any time

Resumable file upload 

1. Download repository.
2. Open README and follow the instruction.
3. Once the transfer completes successfully, check that the file size listed matches the size of the local copy of the file.
   
   

How to download a large file 

If you need to download a large file (>1GB), you will need to use wget or CURL with retries. Copy your file download link from transfer.atlassian.com and use the script as follows:

The following commands will download a file via a HTTP request wget or CURL. You can use any value of retries depending on your internet connection. The optimal value is 250.

$ wget 'https://api.media.atlassian.com/file/0f5b9527-6d46-4d60-85d8-4f2bd164e666 \
/binary?token=token&client=89838f12-f13o-4b52-8f6c-fb55c95dfee6&collection=3547771&dl=true' \
-O filename --tries=250

$ curl 'https://api.media.atlassian.com/file/0f5b9527-6d46-4d60-85d8-4f2bd164e666 \
/binary?token=token&client=89838f12-f13o-4b52-8f6c-fb55c95dfee6&collection=3547771&dl=true' \
-o filename --retry 250

How to get access

If you need to transfer a large file or files to Atlassian, you will need to first create a support request at support.atlassian.com. Once you've created a request, you'll be able to log in to transfer.atlassian.com and upload files against your support issue. 

How to compress data

To minimize bandwidth usage, compress the file into a .ZIP, .GZ, or .TAR.GZ file. (This will also help minimize the opportunity for your connection to drop-out in the middle of a large transfer!)

Linux

In Linux, the tar command works well for this purpose:

$ tar czf ~/jiradb.tar.gz ~/jiradb

Windows or Mac

In Windows or OS X, right clicking on a folder offers a compression mechanism.

How to split data into smaller pieces

Although transfer.atlassian.com has a file size limit 2TB, we recommend limiting files to 1GB in size. This minimizes the window of opportunity for the upload process to fail, and, (in case of a failed upload,) minimizes the time spent re-uploading the same information multiple times.

Linux or Mac

The split command will allow you to break up a file into multiple chunks of a specified size. In the example below, the compressed "jiradb" file is broken into multiple segments <= 1GB in size:

$ split -b 1073741824 ~/jiradb.tar.gz jiradb.tar.gz_
$ ls -l | grep jira
-rw-r--r--    1 user  staff  1073741824 Mar 27 12:03 jiradb.tar.gz_aa
-rw-r--r--    1 user  staff  1073741824 Mar 27 12:03 jiradb.tar.gz_ab
-rw-r--r--    1 user  staff   412536832 Mar 27 12:03 jiradb.tar.gz_ac

The "-b" argument specifies the maximum file size of each segment in bytes. (1073741824 bytes = 1GB) ~/jiradb.tar.gz is the input filename, and jiradb.tar.gz_ is the output filename. By default, the split command appends a two letter code to the end of the specified output filename to indicate the order of the segments. In this case, specifying the "jiradb.tar.gz_" filename resulted in the following sequential segments: jiradb.tar.gz_aa, jiradb.tar.gz_ab, and jiradb.tar.gz_ac. All three of these files will need to be uploaded to transfer.atlassian.com.

Be sure to review your operating system's documentation to be sure you use this tool correctly.

To merge/join the files, you may explore the example below:

$ cat jiradb.tar.gz_* > jiradb.tar.gz
$ tar -cvvzf jiradb.tar.gz .

Windows

Use HJ-Split's split function.