You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
When I reload the hitch daemon (in Ubuntu 16.04 systemd), I get following errors:
Apr 25 19:42:33 localhost systemd[1]: Reloading Hitch TLS unwrapping daemon.
Apr 25 19:42:33 localhost hitch[4035284]: Received SIGHUP: Initiating configuration reload.
Apr 25 19:42:33 localhost hitch[4035284]: 20170425T194233.871567 [4035284] Received SIGHUP: Initiating configuration reload.
Apr 25 19:42:33 localhost systemd[1]: Reloaded Hitch TLS unwrapping daemon.
Apr 25 19:42:33 localhost hitch[4035284]: Config reload failed: Unable to open configuration file '/etc/hitch/hitch.conf': No such file or directory
Apr 25 19:42:33 localhost hitch[4035284]: 20170425T194233.895107 [4035284] Config reload failed: Unable to open configuration file '/etc/hitch/hitch.conf': No such file or directory
After tracing code, I found that the reason was I had turned on the chroot config to /tmp, so the hitch service was actually jailed and unable to see the config file and the certificates outside.
I am wondering if this restriction can be solved, to have both chroot security and reloading mechanism.
The text was updated successfully, but these errors were encountered:
Hitch needs configuration file and certificates. I don't think it is secure to put my certificates together in the jail.
Besides, hitch still wants to chroot to the specified location after reload command. That is, if I specify /hitch as my chroot directory, after the first start, it changes its root to /hitch successfully. But after reload, it wants to chroot to /hitch under current root, that is /hitch/hitch.
Apr 25 20:41:40 localhost systemd[1]: Reloading Hitch TLS unwrapping daemon.
Apr 25 20:41:40 localhost hitch[428268]: Received SIGHUP: Initiating configuration reload.
Apr 25 20:41:40 localhost hitch[428268]: 20170425T204140.130764 [428268] Received SIGHUP: Initiating configuration reload.
Apr 25 20:41:40 localhost hitch[428268]: Config reload failed: Error in configuration file '', line 187: Unable to stat directory '/hitch': No such file or directory'.
Apr 25 20:41:40 localhost hitch[428268]: 20170425T204140.131366 [428268] Config reload failed: Error in configuration file '', line 187: Unable to stat directory '/hitch': No such file or directory'.
Apr 25 20:41:40 localhost systemd[1]: Reloaded Hitch TLS unwrapping daemon.
Not only chroot command will cascade, but also the location of configuration file and certificates. Thus if I use chroot with other than / directory, it is weird to put needed files in the jail if I want multiple reload action.
When I reload the hitch daemon (in Ubuntu 16.04 systemd), I get following errors:
After tracing code, I found that the reason was I had turned on the
chroot
config to/tmp
, so the hitch service was actually jailed and unable to see the config file and the certificates outside.I am wondering if this restriction can be solved, to have both
chroot
security and reloading mechanism.The text was updated successfully, but these errors were encountered: