Dear all, I've got a problem, that's seems to be to do with curl not being able to retrieve a pdf file.
>From the user perspective, this happens: For instance, on this page http://www.ict4e.net/wiki/Tutorials/Dynamic_Range_Compression I can generate pdf for the page to get: http://www.ict4e.net/wiki_source/index.php?title=Special:Book/download/&collection_id=30f6d6e8d8e68492&writer=rl&return_to=Tutorials%2FDynamic+Range+Compression However, when I click the link to get the pdf: http://www.ict4e.net/wiki_source/index.php?title=Special:Book/download/&collection_id=30f6d6e8d8e68492&writer=rl&return_to=Tutorials%2FDynamic+Range+Compression the pdf often doesn't download. More precisely: Sometimes, the pdf doesn't download at all, but sometimes it does. If it downloaded, clicking the back button, and trying to download again, may not work. Typically at the latest by the third download, it no longer works. When the pdf doesn't download, and you click the 'cancel' button on the browser, a php session is left running on the server. It seems that killing all those sessions helps with being able to download again. However, even with no sessions left hanging, the pdf download fails as described above. I've tried this with several browsers (Safari, Firefox, Opera) and on several pages on my wiki, and I get the same behaviour. It happened on v.1.13.2, but have now upgraded to 1.14.0 (and latest version of Collection) and the issue occurs just the same. This is with PHP 5.2.4. I then had a look at the code, at the curl request in function post() in Collection.body.php. Normally when the function is called, it's set to not timeout. So I added a timeout with 10 secs. $timeout = true; $PostFields = array ( 'collection_id' => '30f6d6e8d8e68492', 'writer' => 'rl', 'command' => 'download', ); $wgHTTPTimeout=10; Then, in situations where the download previously stalled, you now get the following error message: Operation timed out after 10000 milliseconds with 40333 out of 261808 bytes received which seems to be typical for this failure: The operation times out with around 40kB read. (Sometimes 0, but most of the time around 40kB it seems.) Of course, where the download works, 10s is more than enough to retrieve the full pdf file. When a failure occurs, curl_getinfo gives: array ( 'url' => 'http://tools.pediapress.com/mw-serve/', 'content_type' => 'application/pdf', 'http_code' => 200, 'header_size' => 211, 'request_size' => 405, 'filetime' => -1, 'ssl_verify_result' => 0, 'redirect_count' => 0, 'total_time' => 10.222578, 'namelookup_time' => 0.002626, 'connect_time' => 0.046523, 'pretransfer_time' => 0.046526, 'size_upload' => 365, 'size_download' => 40333, 'speed_download' => 3945, 'speed_upload' => 35, 'download_content_length' => 261808, 'upload_content_length' => 365, 'starttransfer_time' => 0.046615, 'redirect_time' => 0, ) It seems to me that this must be a server-side issue with the mw- serve. Has anybody else experienced this problem? Any suggestions for further output I could generate to clarify the issue? Any suggestions and help would be greatly appreciated. Many thanks! Bjoern --~--~---------~--~----~------------~-------~--~----~ You received this message because you are subscribed to the Google Groups "mwlib" group. To post to this group, send email to [email protected] To unsubscribe from this group, send email to [email protected] For more options, visit this group at http://groups.google.com/group/mwlib?hl=en -~----------~----~----~----~------~----~------~--~---
