Hi, I'm trying to use "Export to file" to copy a history from one Galaxy instance to another. But rather than giving me the URL link that I can then paste into the second instance, the function is just downloading the history .gz archive directly to my local disk (this is a history that I have 'exported' before, so I guess that because the archive already exists, it doesn't get recreated). But because it's not possible to import from a file, only from a URL, the download is useless - is there a way to force the "Export to file" function to re-supply the download link please?
Hello,
Right now the best way to transfer data is by individual dataset. Download datasets, or download the entire history as a tar file, uncompress it locally, then upload what you need into another Galaxy.
This occurs (direct file download, no link provided) after the download to file function has been used a few times already. There is a limit on how many times an archive will be created. From our testing, the tar file downloaded will be complete if you use the Google Chrome browser.
The link functionality was not working exactly as expected when we all tested here last week (the import portion). We consider addressing this a priority. Any updates that our team has right now we'll post to this thread, otherwise please follow progress on released updates here: https://wiki.galaxyproject.org/DevNewsBriefs
Tickets to address enhancements for this function are in Trello, this is one of them: https://trello.com/c/qCfAWeYU
Best, Jen, Galaxy team
Hi Mark,
if you can download the file, there must be a link or? Right click on the download button and copy the link.
Cheers,
Bjoern
Thanks Bjoern, but the "Export to file" link isn't to a file location, but instead to a javascript function (which presumably then either launches an archiving function or, if the archive already exists, initiates a download). So if I right-click and copy the link, all I get is "javascript:void(0);".
Cheers
Mark
This must be because of your browser; I tested in two different Galaxy instances and each time I get a regular link that I can either copy (through right-click or, of course by selecting the text and copy) or click to get the .tar.gz file.
As far as I can tell though, it is not possible to use a local stored file to upload a history (i.e. using 'file:///full/path/to/history.tar.gz' doesn't seem to work even though I do not get an error).
Which browser did you use? Can you try a different one?
It's not a browser issue - I get the same with Opera, Safari, Chrome and Firefox, on Mac and Windows, and with at least two separate Galaxy instances. Just to confirm, this is using the 'cog' icon at the top of the History panel, and selecting 'Export to file", about four up from the bottom of the list.
When I do this the first time on any given history, I get a blue box saying "Exporting History '[History name]'. Use this link to download the archive or import it to another Galaxy server: [link here]". But if I click it again later (if I need to go back and re-export the history and didn't record that link the first time) it immediately downloads the archive to my local computer (which as you point out, can't then be used for an upload). Right clicking to save the link gives nothing but "javascript:void(0);".
Are either of the instances you tested it on public? If so, please could you let me know what they are so I can see if it works for me too on them? I don't know if it could be a Galaxy version issue, where it has changed from javascript to a direct link, but that might explain why it's not reproducible. One of the ones I tried that doesn't work for me was galaxy-qld.genome.edu.au (test example at http://galaxy-qld.genome.edu.au/u/markc/h/unnamed-history).
Cheers
Mark