chroot sftp stopped working today

[ultracaffeinated]

Plesk 12.0.18, no yum updates, just stopped working. System accounts work, plesk created accounts with their shell set to /bin/bash (chrooted) fail. I'm getting this error in /var/log/messages:

Feb 25 01:08:31 ip-XXX-XXX-XXX-XXX systemd: Created slice user-10002.slice.
Feb 25 01:08:31 ip-XXX-XXX-XXX-XXX systemd: Starting user-10002.slice.
Feb 25 01:08:31 ip-XXX-XXX-XXX-XXX systemd-logind: New session 904 of user ftplogin.
Feb 25 01:08:31 ip-XXX-XXX-XXX-XXX systemd: Started Session 904 of user ftplogin.
Feb 25 01:08:31 ip-XXX-XXX-XXX-XXX systemd: Starting Session 904 of user ftplogin.
Feb 25 01:08:31 ip-XXX-XXX-XXX-XXX plesk-chrootsh[30686]: execv("/bin/bash") failed
Feb 25 01:08:31 ip-XXX-XXX-XXX-XXX plesk-chrootsh[30686]: system error: Permission denied
Feb 25 01:08:31 ip-XXX-XXX-XXX-XXX systemd-logind: Removed session 904.
Feb 25 01:08:31 ip-XXX-XXX-XXX-XXX systemd: Removed slice user-10002.slice.
Feb 25 01:08:31 ip-XXX-XXX-XXX-XXX systemd: Stopping user-10002.slice.

I've run the plesk repair tool and rebuilt all the configs with httpdmng. I've also followed these directions: http://kb.odin.com/en/125519 to no avail (they do get rid of the PAM faulty module message but don't restore chroot. Any other ideas? CentOS 7 with SELlinux set to permissive for now.

 

[ultracaffeinated]

I adjusted the subscription to have a different shell, synced them all, then ran the bootstrapper repair:

/usr/local/psa/bootstrapper/pp12.0.18-bootstrapper/bootstrapper.sh repair

Then I switched it back to /bin/bash (chrooted), synced, and it started working. Running the bootstrapper repair originally did nothing.