In addition to below,
It looks like this might be this issue, previously resolved with mariadb
but possibly not with packetfence-dhcpd.
https://github.com/inverse-inc/packetfence/issues/2792
Our short term workaround was to just turn off the broken systemd process
notification by changing the service type to "simple", executing a systemd
daemon-reload and restarting the packetfence-dhcpd.service.
Seems to be good for now.
cheers,
Ian
On Wed, Feb 21, 2018 at 10:28 AM, Ian MacDonald <i...@netstatz.com> wrote:
> Hello,
>
> Our PF instance(s) run in an out-of-band configuration, providing a
> captive portal to hostapd/CoA enabled switches using connection profiles
> with SSID filters and an email source for activation.
>
> We are not clustering, but instead have a qa instance where we stage and
> test upgrades and configuration changes.
>
> Since at least pf 7.3.0 we have noticed that the dhcpd service is
> constantly restarting every couple of minutes. We just completed our
> upgrade to 7.4.0 and we still continue to see this dhcpd service restarting
> every couple of minutes. This service provides connectivity to the
> registration and isolation VLANs by default.
>
> In our logs we see the lines captured below; I have included a capture
> from both our instances which behave the same way.
>
> Out best guess is that for some reason the dhcpd service monitoring is not
> quite functioning correctly, and restarting the service on a periodic
> basis. The "Failed to start" message even though it has started, seems to
> be an indication of perhaps some issue with the startup/systemd integration
> where it is not properly detecting the service start and simply restarting
> after the holdoff timer has expired.
>
> The net impact is that we believe this is causing some inconsistencies in
> portal registration / authorization timing that we may have previously
> worked around by increasing our redirect delay on activation.
>
> Now that we are on the latest version, we are posting here to see if there
> is a known quick fix while we continue to poke around with the dhcpd
> service startup configuration.
>
> cheers,
> Ian
>
> Instance 1:
> Feb 21 10:05:18 pf2 systemd[1]: packetfence-dhcpd.service start operation
> timed out. Terminating.
> Feb 21 10:05:18 pf2 systemd[1]: Failed to start PacketFence DHCPv4 Server
> Daemon.
> Feb 21 10:05:18 pf2 systemd[1]: Unit packetfence-dhcpd.service entered
> failed state.
> Feb 21 10:05:18 pf2 systemd[1]: packetfence-dhcpd.service holdoff time
> over, scheduling restart.
> Feb 21 10:05:18 pf2 systemd[1]: Stopping PacketFence DHCPv4 Server
> Daemon...
> Feb 21 10:05:18 pf2 systemd[1]: Starting PacketFence DHCPv4 Server
> Daemon...
> Feb 21 10:05:21 pf2 pfcmd[1952]: service|command
> Feb 21 10:05:21 pf2 pfcmd[1952]: dhcpd|config generated
>
> Instance 2:
> Feb 21 10:01:11 pf4 systemd[1]: packetfence-dhcpd.service start operation
> timed out. Terminating.
> Feb 21 10:01:11 pf4 systemd[1]: Failed to start PacketFence DHCPv4 Server
> Daemon.
> Feb 21 10:01:11 pf4 systemd[1]: Unit packetfence-dhcpd.service entered
> failed state.
> Feb 21 10:01:11 pf4 systemd[1]: packetfence-dhcpd.service holdoff time
> over, scheduling restart.
> Feb 21 10:01:11 pf4 systemd[1]: Stopping PacketFence DHCPv4 Server
> Daemon...
> Feb 21 10:01:11 pf4 systemd[1]: Starting PacketFence DHCPv4 Server
> Daemon...
> Feb 21 10:01:15 pf4 pfcmd[16671]: service|command
> Feb 21 10:01:15 pf4 pfcmd[16671]: dhcpd|config generated
>
>
------------------------------------------------------------------------------
Check out the vibrant tech community on one of the world's most
engaging tech sites, Slashdot.org! http://sdm.link/slashdot
_______________________________________________
PacketFence-users mailing list
PacketFence-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/packetfence-users
