Re: [Fwd: [Duplicity-talk] 0.5.18 weird GPG error ([key_I_dont_mention] unusable public key)]

[Kenneth Loafman]

This patch will fix the waitpid() issue, but not the bad key issue.  I'm
going to have to chase that one a bit.  If its trusted in his keyring
then it should work.

Edgar Soldin wrote:
> Ken,
> 
> did you read this? I also experienced the
>  OSError: [Errno 10] No child processes
> and am therefore still using 0.5.17 ... Maybe you could consider modifying 
> (or removing?) the 0.5.18 patch?
> 
> thanks ede
> 
> 
> -------- Original Message --------
> Subject: [Duplicity-talk] 0.5.18 weird GPG error ([key_I_dont_mention]        
> unusable public key)
> Date: Wed, 8 Jul 2009 09:38:38 -0400
> From: Andrew Kohlsmith (Mailing List Account) <address@hidden>
> Reply-To: address@hidden,     Discussion of the backup program duplicity      
> <address@hidden>
> To: address@hidden
> 
> Good morning,
> 
> I am running into an unusual problem with duplicity. It fails when doing a 
> full backup, stating that a key has no indication that it really belongs to 
> the owner, followed by an "encryption failed: unusable public key" error.
> 
> The problem is that the fingerprint it's giving me as being the problem is 
> NOT 
> one that I have specified!
> 
> My command line:
> # duplicity --full-if-older-than 26W --archive-dir 
> /var/data/duplicity/archivedir/test --encrypt-key KEYONE --encrypt-key KEYTWO 
> --encrypt-key KEYTHREE /var/data/cvs file:///var/data/duplicity/test          
>                                    
> 
> obviously KEYONE, KEYTWO and KEYTHREE are my GPG keys. (I encrypt to my 
> personal key, a backup key, and a company key.)
> 
> The backup fails almost immediately with the following message:
> 
> Last full backup date: none                                                   
>             
> Last full backup is too old, forcing full backup
> Exception in thread wait26030:
> Traceback (most recent call last):
>   File "/usr/lib/python2.3/threading.py", line 436, in __bootstrap
>     self.run()
>   File "/usr/lib/python2.3/threading.py", line 416, in run
>     self.__target(*self.__args, **self.__kwargs)
>   File "/usr/lib/python2.3/site-packages/duplicity/GnuPGInterface.py", line 
> 668, in threaded_waitpid
>     process.returned = os.waitpid(process.pid, 0)[1]
> OSError: [Errno 10] No child processes
> 
> Exception in thread wait26027:
> Traceback (most recent call last):
>   File "/usr/lib/python2.3/threading.py", line 436, in __bootstrap
>     self.run()
>   File "/usr/lib/python2.3/threading.py", line 416, in run
>     self.__target(*self.__args, **self.__kwargs)
>   File "/usr/lib/python2.3/site-packages/duplicity/GnuPGInterface.py", line 
> 668, in threaded_waitpid
>     process.returned = os.waitpid(process.pid, 0)[1]
> OSError: [Errno 10] No child processes
> 
> GPGError: GPG Failed, see log below:
> ===== Begin GnuPG log =====
> gpg: FF873F7F: There is no indication that this key really belongs to the 
> owner
> gpg: [stdin]: encryption failed: unusable public key
> ===== End GnuPG log =====
> 
> Now the curious part is that that key, FF873F7F, is NOT one I have specified 
> on 
> my command line.  After some initial debugging, I have found the source of 
> that key.  It's a "sub" key on my personal signing/encryption key:
> 
> # gpg --list-key EAF7ACB0
> pub  1024D/EAF7ACB0 2001-08-27 Andrew Kohlsmith <address@hidden>
> uid                            Andrew Kohlsmith <address@hidden>
> uid                            Andrew Kohlsmith (mailing lists account) 
> <address@hidden>
> sub  2048g/FF873F7F 2001-08-27
> 
> Now I don't have any problem with this key with anything else; why is 
> duplicity unhappy?
> 
> -A.
> 
> 
> 
> _______________________________________________
> Duplicity-talk mailing list
> address@hidden
> http://lists.nongnu.org/mailman/listinfo/duplicity-talk
> 
> 
> _______________________________________________
> Duplicity-talk mailing list
> address@hidden
> http://lists.nongnu.org/mailman/listinfo/duplicity-talk
> 

=== modified file 'duplicity/GnuPGInterface.py'
--- duplicity/GnuPGInterface.py 2009-05-13 17:15:44 +0000
+++ duplicity/GnuPGInterface.py 2009-07-10 15:46:11 +0000
@@ -665,7 +665,11 @@
     will allow us to reap child processes as soon as possible,
     thus freeing resources quickly.
     """
-    process.returned = os.waitpid(process.pid, 0)[1]
+    try:
+        process.returned = os.waitpid(process.pid, 0)[1]
+    except:
+        log.Debug("GPG process %d terminated before wait()" % process.pid)
+        process.returned = 0
 
 
 def _run_doctests():

signature.asc
Description: OpenPGP digital signature