On Fri 2017-02-03 09:02:47 -0500, Yuri D'Elia wrote: > On Fri, Feb 03 2017, Michael Biebl wrote: >>> When gpg is used via command line, the agent is started automatically >>> and then it's left sitting there. But for a system without seats, the >>> agent doesn't make sense. It shouldn't be running. >> >> That doesn't make sense. Even if you stop the sockets, once you use gpg, >> gpg-agent will be auto-started as well, just using a different mechanism. > > It this also true also when --no-autostart is in use?
In this case, gpg-agent will not be automatically started, but if it needs any information from the secret keyring (which is the only reason it ever tries to auto-launch gpg-agent in the first place), it will fail. What are you doing with gpg? if whatever you're doing needs the secret keyring, the agent will be launched. if it doesn't need the secret keyring, the agent will not be launched. With gpg-agent as a systemd user service, when the user completely logs out of the system, any services launched (socket-activated or otherwise) will be stopped automatically. OTOH, if gpg-agent is auto-launched by gpg (not managed as a systemd user service), there is no robust standard way to ensure that the agent gets terminated at logout. Yuri, it sounds like you've got a particular scenario that you don't like, and you're trying lots of things to change it, but i don't understand what the scenario is. Can you try to distill the problem you're seeing into a repeatable pattern? I have a minimal server running debian testing. when i log into it with ssh, and use gpg with secret key material, the gpg-agent is spawned by systemd's socket activation, and remains supervised by systemd. When i log out, the gpg-agent process gets terminated cleanly by systemd. What's the problem? --dkg
signature.asc
Description: PGP signature