gpg crash on restore

Asked by rafael

Hi!

I have rolled out GPG/token duplicity backups at work. One of our server's HD has crashed, so I am in the process of restoring it.

I have a new system where I have setup GPG and it works,I can successfully run card-status, encrypt, decrypt, etc. All with the same token the backup was encrypted for.

I can even decrypt the pgp files that duplicity created (manually, as test).

However, duplicity restore crashes with:

pcsc_connect failed: sharing violation (0x8010000b), in the very same terminal, with the very same user etc.

I also set the PASSPHRASE variable to avoid any kind of AGENT issues.

I have already try on two ubuntu systems. 14.x 15.x with the same result. in x86 and armf. All with the same result.

Any idea what could be going on? Any idea how could I debug it?

Regards,
Rafael

Question information

Language:
English Edit question
Status:
Answered
For:
Duplicity Edit question
Assignee:
No assignee Edit question
Last query:
Last reply:
Revision history for this message
edso (ed.so) said :
#1

On 26.02.2016 07:47, rafael wrote:
> New question #286882 on Duplicity:
> https://answers.launchpad.net/duplicity/+question/286882
>
> Hi!
>
> I have rolled out GPG/token duplicity backups at work. One of our server's HD has crashed, so I am in the process of restoring it.
>
> I have a new system where I have setup GPG and it works,I can successfully run card-status, encrypt, decrypt, etc. All with the same token the backup was encrypted for.
>
> I can even decrypt the pgp files that duplicity created (manually, as test).
>
> However, duplicity restore crashes with:
>
> pcsc_connect failed: sharing violation (0x8010000b), in the very same terminal, with the very same user etc.
>
> I also set the PASSPHRASE variable to avoid any kind of AGENT issues.
>
> I have already try on two ubuntu systems. 14.x 15.x with the same result. in x86 and armf. All with the same result.
>
> Any idea what could be going on? Any idea how could I debug it?
>

did you try
 http://lists.gnupg.org/pipermail/gnupg-users/2007-July/031499.html
already?

wrt. to debugging. running duplicity< w/ max. verbosity might give you a more detailed gpg output.

..ede/duply.net

Revision history for this message
rafael (rvalle) said :
#2

Yes,

Like many others, I experience errors with GPG. For me, killing the gnome-keyring, restarting pcscd daemon and/or reinserting the card makes thing work again.

I normally issue gpg --card-status to check that everything is ok, even I decipher a hello world file.

It is not a problem for me to the GPG to work. I achieve that consistently despite the fact that it is a hassle.

The problem here is that duplicity seems to use another configuration/environment unrelated to the gpg user environment setup.

When gpg is perfectly working duplicity consistently fails.

I am going to try to generate max verbosity.

R.

Revision history for this message
edso (ed.so) said :
#3

On 27.02.2016 10:46, rafael wrote:
> Question #286882 on Duplicity changed:
> https://answers.launchpad.net/duplicity/+question/286882
>
> rafael posted a new comment:
> Yes,
>
> Like many others, I experience errors with GPG. For me, killing the
> gnome-keyring, restarting pcscd daemon and/or reinserting the card makes
> thing work again.
>
> I normally issue gpg --card-status to check that everything is ok, even
> I decipher a hello world file.
>
> It is not a problem for me to the GPG to work. I achieve that
> consistently despite the fact that it is a hassle.
>
> The problem here is that duplicity seems to use another
> configuration/environment unrelated to the gpg user environment setup.
>
> When gpg is perfectly working duplicity consistently fails.
>
> I am going to try to generate max verbosity.
>

did you try duply already? it has a gpg test, which tries to mimic the command line enc/decryption within duplicity. maybe this helps you to narrow it down?

..ede/duply.net

Can you help with this problem?

Provide an answer of your own, or ask rafael for more information if necessary.

To post a message you must log in.