This discussion has been locked.
You can no longer post new replies to this discussion. If you have a question you can start a new discussion

UTM fails to power off after shutdown or gets stuck

We have a small SG 115 UTM device at the office. We also have a UPS at the office which keeps the few servers we have running long enough for them to perform a graceful shutdown. In theory, I want the UTM to perform a graceful shutdown as well, but I'm having some issues with this.

A simple command is sent to the UTM via PUTTY: shutdown now

The command is clearly received by the UTM as the process is initiated. However, the UTM never really powers down. Well, if I connect a screen to the UTM, I get the following (terribly sorry for the poor quality): https://imgur.com/HTOT4Me

The last lines seem to read as follows:

Stopping Logging | done
Master Resource Control: runlevel 1 has been
Skipping services in runlevel 1: | utm-??? (I honestly can't make it out from the photo; yes, I know it's bad)

Additionally, looking into the system log, this is all I could find that seemed to be related to the actual shutdown:

2019:06:04-16:10:59 firewall postgres[3875]: [3-1] LOG:  received fast shutdown request<134>Jun  4 16:10:59 postgres[3875]: [4-1] LOG:  aborting any active transactions
2019:06:04-16:10:59 firewall postgres[22476]: [3-1] FATAL:  terminating connection due to administrator command
2019:06:04-16:10:59 firewall postgres[22899]: [3-1] FATAL:  terminating connection due to administrator command
2019:06:04-16:10:59 firewall postgres[22297]: [3-1] FATAL:  terminating connection due to administrator command
2019:06:04-16:10:59 firewall postgres[22478]: [3-1] FATAL:  terminating connection due to administrator command
2019:06:04-16:10:59 firewall postgres[22263]: [3-1] FATAL:  terminating connection due to administrator command
2019:06:04-16:10:59 firewall postgres[22329]: [3-1] FATAL:  terminating connection due to administrator command
2019:06:04-16:10:59 firewall postgres[22254]: [3-1] FATAL:  terminating connection due to administrator command
2019:06:04-16:10:59 firewall postgres[22260]: [3-1] FATAL:  terminating connection due to administrator command
2019:06:04-16:10:59 firewall postgres[3880]: [3-1] LOG:  autovacuum launcher shutting down
2019:06:04-16:10:59 firewall postgres[22252]: [3-1] FATAL:  terminating connection due to administrator command
2019:06:04-16:10:59 firewall postgres[5392]: [3-1] FATAL:  terminating connection due to administrator command
2019:06:04-16:10:59 firewall postgres[4704]: [3-1] FATAL:  terminating connection due to administrator command
2019:06:04-16:10:59 firewall postgres[3877]: [2-1] LOG:  shutting down<134>Jun  4 16:10:59 postgres[3877]: [3-1] LOG:  database system is shut down
2019:06:04-16:11:01 firewall syslog-ng[4902]: syslog-ng shutting down; version='3.4.7'

What could be going on? I don't know if the UTM is safe to power off at this state or not - I'd prefer it to power off completely automatically. I haven't tried a manual reboot yet (will do so tomorrow, when I'm at the office), but I do know that the UTM has no trouble rebooting when installing updates.



This thread was automatically locked due to age.
Parents
  • Found the reason, by testing the UTM on a VM (so that I could play around with the commands as much as I please without impacting the office).

    It seems the shutdown command doesn't, in fact, turn the system off at the end. Event with the -P parameter the end result is just as I described above. Instead it seems better to just use the halt command. This one works "as advertised". If someone wants to use shutdown to power off, then the additional -h flag is required (for some reason -P does nothing).

  • In fact, Mateusz, no graceful shutdown is needed.

    Cheers -Bob

     
    Sophos UTM Community Moderator
    Sophos Certified Architect - UTM
    Sophos Certified Engineer - XG
    Gold Solution Partner since 2005
    MediaSoft, Inc. USA
  • Any chance you could elaborate on this? Given the choice to shutdown gracefully and not, I'd say the former is always preferable, no?

  • In 16 years of being around this platform, I've never seen a customer have an issue with a "graceless" shutdown, Mateusz.  The only threads here about graceful shutdown were folks wanting to have everything shut down gracefully.  I don't have perfect memory, but I don't recall a single one where someone had a problem with their UTM and wanted to avoid a future problem.  That said, I do agree that guaranteeing a graceful shutdown might make a difference someday.

    Cheers - Bob

     
    Sophos UTM Community Moderator
    Sophos Certified Architect - UTM
    Sophos Certified Engineer - XG
    Gold Solution Partner since 2005
    MediaSoft, Inc. USA
Reply
  • In 16 years of being around this platform, I've never seen a customer have an issue with a "graceless" shutdown, Mateusz.  The only threads here about graceful shutdown were folks wanting to have everything shut down gracefully.  I don't have perfect memory, but I don't recall a single one where someone had a problem with their UTM and wanted to avoid a future problem.  That said, I do agree that guaranteeing a graceful shutdown might make a difference someday.

    Cheers - Bob

     
    Sophos UTM Community Moderator
    Sophos Certified Architect - UTM
    Sophos Certified Engineer - XG
    Gold Solution Partner since 2005
    MediaSoft, Inc. USA
Children
  • I'll add I have seen occasional issues with UTM and a less than graceful shutdown; not very often (maybe 2 or 3 times in the 15 years I've resold and managed the platform) though.  Typical results are either corrupted reporting data or corrupt email queues.  It's not something I really worry about though.

    CTO, Convergent Information Security Solutions, LLC

    https://www.convergesecurity.com

    Sophos Platinum Partner

    --------------------------------------

    Advice given as posted on this forum does not construe a support relationship or other relationship with Convergent Information Security Solutions, LLC or its subsidiaries.  Use the advice given at your own risk.

  • Thanks, Bruce - I was hoping that another one of the "old-timers" would chime in.  That's the one thing I can think of.  I've never seen that happen, but I have seen Up2Dates break a PostgreSQL database many times in my customer base and here in the Community - not every Up2Date is graceful!  A database rebuild fixes that, losing data.

    Cheers - Bob

     
    Sophos UTM Community Moderator
    Sophos Certified Architect - UTM
    Sophos Certified Engineer - XG
    Gold Solution Partner since 2005
    MediaSoft, Inc. USA