Problems syncing large files on WebDAV, NS_ERROR_FAILURE: Couldn't decrypt string(Report: 952020678)

danmerchant

I have recently added some large PDF's (approx. 120MB) to my library hosted on my home WebDAV server. At home I was able to add the files and sync them. When I try to access them on another computer/mobile app I get the following error messages:

[JavaScript Error: "NS_ERROR_FAILURE: Couldn't decrypt string" {file: "resource://gre/components/crypto-SDR.js" line: 146}]

[JavaScript Error: "Download request 1/5Q2RCZZS failed"]

[JavaScript Error: "Your WebDAV server returned an HTTP 0 error for a GET request.

I have tried downloading the PDF manually and placing it in to the appropriate storage folder on the local machine but I am still encountering this error. Previously, I had an issue with PHP restricting the maximum file size for transfers but this is resolved now. There are no errors in the Apache logs regarding this issue. It looks like an encryption issue to me but I am really not sure. Any help with this issue is much appreciated.

Regards

Dan

dstillman

In the Sync prefs, are you able to verify the server from that computer?

Are you connecting via a proxy?

dstillman

For the "Couldn't decrypt string" error, you can try deleting key8.db and cert8.db from your Zotero profile directory, restarting Zotero, and setting up syncing again.

But an HTTP 0 error means that the request itself is failing or being interrupted, and if you're getting that far then previous requests to the server are working. So that would really point to Apache (or something in between, if you've overridden the certificate protections) crashing or otherwise terminating the connection when you try to download that specific file.

danmerchant

Yes I am connecting via a proxy, I am able to Verify Server sucessfully. I have tried deleting key8.db and cert8.db and restarting but I am still getting the same error on sync.

Interestingly, I have just downloaded the zip file in question manually through a web browser and I am getting a data error when trying to extract the file from the zip. It would appear that the zip file is corrupted.

dstillman

That still likely just means that the connection is being interrupted — possibly by your proxy server — and your browser is leaving behind a partial file, which obviously wouldn't extract. Zotero is just properly reporting the interrupted connection before it tries to process the file.

danmerchant

The file size on the WebDAV and the downloaded file size match, also the MD5 checksums match. I will try removing the file from my library on my home computer when I get home and try re-adding it to see if that fixes the problem.

dstillman

Is your proxy doing some sort of virus scanning? The internal integrity of the ZIP file obviously shouldn't cause the connection to fail.

danmerchant

I think that the proxy probably is doing some sort of virus scanning. Often when I download things, they hang on 100% for quite some time. I assume that is the proxy virus checker. Interestingly, this PDF has the message "This file claims compliance with the PDF/A standard and hasbeen opened read-only to prevent modification." I will test things out at home when I am not behind this proxy and see if it is working then.

danmerchant

Is it possible that as the download is large and then being delayed at the end by the proxy virus checker that it could be exceeding one of the timeout values? If so, could you advise on which one could be increased in the Config Editor?

danmerchant

I searched the Config Editor for "timeout" preferences. I doubled all the values and this appears to have rectified the issue.