[Bug 1793572] Re: systemd-logind crashes with NIS users

[Dan Streetman]

please reopen if this is still an issue

** Changed in: systemd (Ubuntu)
       Status: New => Invalid

-- 
You received this bug notification because you are a member of Ubuntu
Foundations Bugs, which is subscribed to systemd in Ubuntu.
https://bugs.launchpad.net/bugs/1793572

Title:
  systemd-logind crashes with NIS users

Status in systemd package in Ubuntu:
  Invalid

Bug description:
  Description:    Ubuntu 18.04.1 LTS
  Release:        18.04

  
  In a fresh Ubuntu 18.04.1 install, after having set up NIS for connecting to my company's network, I noticed several problems happening when I tried to log in with my NIS user account.

  Initially logged in with a local user, I set up NIS client, and
  confirmed that NIS queries were OK, and then I tried to "ssh
  my_nis_acct at localhost", which hanged for a while, and them my X
  session crashed.

  Attempts to log in on the text console produced the following
  behavior: after typing in the password and getting some session-
  related messages, the login process hanged for 25 seconds, during
  which the machine became pretty much unusable, but then the login
  completed.  However, a couple of minutes later, the X session of the
  local user (which was left open at tty1) suddenly died.  That
  situation was consistently reproducible.

  Googling words related to the symptoms (most especially the 25-second hangs), I came across a bug report for Fedora, stating that the culprit would be nss_nis (cf. 
  https://bugzilla.redhat.com/show_bug.cgi?id=1576558 and https://bodhi.fedoraproject.org/updates/FEDORA-2018-78f95660f9).

  As an workaround, I disabled pam_systemd.so in every PAM module that
  referred to it.  Now I could login freely with NIS-based accounts, but
  lost the facilities of systemd-logind.

  After having has an idea about who the culprit was and managing to
  restore login for NIS-based users, I indeed saw messages in
  /var/log/syslog that showed at the time when systemd-logind crashed,
  indeed consistently preceded by another indicating something wrong
  related to some NIS query, such as shown in the excerpt below.

  -----------------------------
  Sep 15 07:51:36 my-machine systemd-logind[26803]: do_ypcall: clnt_call: RPC: Unable to send; errno = Operation not permitted
  Sep 15 07:54:36 my-machine systemd[1]: systemd-logind.service: Watchdog timeout (limit 3min)!
  Sep 15 07:54:36 my-machine systemd[1]: systemd-logind.service: Killing process 26803 (systemd-logind) with signal SIGABRT.
  Sep 15 07:54:37 my-machine systemd[1]: systemd-logind.service: Main process exited, code=dumped, status=6/ABRT
  Sep 15 07:54:37 my-machine systemd[1]: systemd-logind.service: Failed with result 'watchdog'.
  Sep 15 07:54:37 my-machine systemd[1]: systemd-logind.service: Service has no hold-off time, scheduling restart.
  Sep 15 07:54:37 my-machine systemd[1]: systemd-logind.service: Scheduled restart job, restart counter is at 2.
  Sep 15 08:10:29 my-machine systemd-logind[1804]: do_ypcall: clnt_call: RPC: Unable to send; errno = Operation not permitted
  Sep 15 08:13:06 my-machine systemd[1]: systemd-logind.service: Watchdog timeout (limit 3min)!
  Sep 15 08:13:06 my-machine systemd[1]: systemd-logind.service: Killing process 1804 (systemd-logind) with signal SIGABRT.
  Sep 15 08:13:07 my-machine systemd[1]: systemd-logind.service: Main process exited, code=dumped, status=6/ABRT
  Sep 15 08:13:07 my-machine systemd[1]: systemd-logind.service: Failed with result 'watchdog'.
  Sep 15 08:13:07 my-machine systemd[1]: systemd-logind.service: Service has no hold-off time, scheduling restart.
  Sep 15 08:13:07 my-machine systemd[1]: systemd-logind.service: Scheduled restart job, restart counter is at 1.
  Sep 15 08:13:38 my-machine systemd-logind[7153]: do_ypcall: clnt_call: RPC: Unable to send; errno = Operation not permitted
  Sep 15 08:16:07 my-machine systemd[1]: systemd-logind.service: Watchdog timeout (limit 3min)!
  Sep 15 08:16:07 my-machine systemd[1]: systemd-logind.service: Killing process 7153 (systemd-logind) with signal SIGABRT.
  Sep 15 08:16:07 my-machine systemd[1]: systemd-logind.service: Main process exited, code=dumped, status=6/ABRT
  Sep 15 08:16:07 my-machine systemd[1]: systemd-logind.service: Failed with result 'watchdog'.
  Sep 15 08:16:07 my-machine systemd[1]: systemd-logind.service: Service has no hold-off time, scheduling restart.
  Sep 15 08:16:07 my-machine systemd[1]: systemd-logind.service: Scheduled restart job, restart counter is at 2.
  --------------------

  I wonder if the fix that was applied in Fedora might work for Ubuntu
  as well.

To manage notifications about this bug go to:
https://bugs.launchpad.net/ubuntu/+source/systemd/+bug/1793572/+subscriptions