http://bugzilla.suse.com/show_bug.cgi?id=1051035 http://bugzilla.suse.com/show_bug.cgi?id=1051035#c3 Ralf Friedl <Ralf.Friedl@online.de> changed: What |Removed |Added ---------------------------------------------------------------------------- Flags|needinfo?(Ralf.Friedl@onlin | |e.de) | --- Comment #3 from Ralf Friedl <Ralf.Friedl@online.de> --- (In reply to Michael Andres from comment #2)
@Ralf Please confirm that you agents do not vanish, even after a couple of minutes. In this case we should pass the bug to the gpg maintainer. I only started to investigate because I noticed around 50 gpg-agent processes shown with "zypper ps -s" after updating. At first I thought strange, but they may be needed, then I noticed the arguments with the temporary directories indicating zypper, so I killed them.
But I have a machine where I didn't kill them. It seems the oldest processes are from May, which should qualify for more than a few minutes. Note that I don't use gpg-agent, which is why I have no problem using killall gpg-agent. Maybe gpg doesn't start a new agent if it finds one running, although that would also be wrong according to your comment #1 it is separate from the user keyring. # for i in $(pidof gpg-agent); do echo /proc/$i; done | xargs ls -lrtd dr-xr-xr-x 9 root root 0 May 20 15:17 /proc/6906 dr-xr-xr-x 9 root root 0 May 20 15:30 /proc/7747 dr-xr-xr-x 9 root root 0 May 20 18:05 /proc/11915 dr-xr-xr-x 9 root root 0 May 22 14:30 /proc/6563 dr-xr-xr-x 9 root root 0 May 22 14:30 /proc/6549 dr-xr-xr-x 9 root root 0 May 22 15:30 /proc/8254 dr-xr-xr-x 9 root root 0 May 30 15:30 /proc/8551 dr-xr-xr-x 9 root root 0 May 30 15:30 /proc/8538 dr-xr-xr-x 9 root root 0 May 30 23:33 /proc/28076 dr-xr-xr-x 9 root root 0 May 30 23:33 /proc/28095 dr-xr-xr-x 9 root root 0 Jun 21 19:26 /proc/23659 dr-xr-xr-x 9 root root 0 Jun 21 19:26 /proc/23667 dr-xr-xr-x 9 root root 0 Jun 23 18:57 /proc/23922 dr-xr-xr-x 9 root root 0 Jun 23 18:57 /proc/23932 dr-xr-xr-x 9 root root 0 Jun 23 18:58 /proc/24065 dr-xr-xr-x 9 root root 0 Jul 11 14:55 /proc/2336 dr-xr-xr-x 9 root root 0 Jul 11 14:55 /proc/2344 dr-xr-xr-x 9 root root 0 Jul 11 14:56 /proc/2554 dr-xr-xr-x 9 root root 0 Jul 11 15:13 /proc/10997 dr-xr-xr-x 9 root root 0 Jul 11 15:13 /proc/11044 dr-xr-xr-x 9 root root 0 Jul 11 15:14 /proc/11174 dr-xr-xr-x 9 root root 0 Jul 17 10:39 /proc/15588 dr-xr-xr-x 9 root root 0 Jul 17 10:39 /proc/15596 dr-xr-xr-x 9 root root 0 Jul 17 10:40 /proc/15919 dr-xr-xr-x 9 root root 0 Jul 17 10:42 /proc/16122 dr-xr-xr-x 9 root root 0 Jul 18 11:58 /proc/8693 You are correct that gpg-agent is started by gpg, so the maintainer should know that the documentation says that --use-standard-socket has no effect. Actually, gpg-agent says the same: write(2, "gpg-agent[20560]: WARNING: \"--use-standard-socket", 49) = 49 write(2, "\" is an obsolete option - it has no effect\n", 43) = 43 gpg only launches gpg-agent after it has read the key from the command line, and the key is not imported into the agent. Than about once a minute the agent forks a process that connects to the agent, does "GETINFO pid" and "BYE". -- You are receiving this mail because: You are on the CC list for the bug.