There is one warning in the log during restart of keepalived.
# journalctl -f | grep keepalived
Oct 09 15:54:05 nadc1-pfence-01 sudo[152287]: root : TTY=pts/1 ;
PWD=/root ; USER=root ; COMMAND=/bin/systemctl restart
packetfence-keepalived
Oct 09 15:54:09 nadc1-pfence-01 packetfence[152297]: -e(152297) INFO:
main, -e, 1 (pf::services::manager::keepalived::generateConfig)
Oct 09 15:54:09 nadc1-pfence-01 Keepalived[152324]: WARNING - default
user 'keepalived_script' for script execution does not exist - please
create.
Oct 09 15:54:09 nadc1-pfence-01 Keepalived[152324]: Opening file
'/usr/local/pf/var/conf/keepalived.conf'.
Oct 09 15:54:09 nadc1-pfence-01 Keepalived_vrrp[152328]: Opening file
'/usr/local/pf/var/conf/keepalived.conf'.
Oct 09 15:54:09 nadc1-pfence-01 packetfence[152108]: pfcmd.pl(152108)
INFO: Daemon keepalived took 3.692 seconds to start.
(pf::services::manager::restartService)
Oct 09 15:54:09 nadc1-pfence-01 Keepalived_healthcheckers[152327]:
Opening file '/usr/local/pf/var/conf/keepalived.conf'.
Oct 09 15:54:09 nadc1-pfence-01 sudo[152333]: root : TTY=pts/1 ;
PWD=/root ; USER=root ; COMMAND=/bin/systemctl show -p MainPID
packetfence-keepalived
Here is the keepalived.conf
# This file is generated from a template at
/usr/local/pf/conf/keepalived.conf
# Any changes made to this file will be lost on restart
global_defs {
notification_email {
jlin...@jerviswebb.com
}
notification_email_from packetfe...@daifukuna.com
smtp_server 10.22.0.92
smtp_connect_timeout 30
router_id PacketFence-nadc1-pfence-01
}
vrrp_track_process radius_load_balancer {
process /usr/sbin/freeradius -d /usr/local/pf/raddb -n load_balancer -fm
full_command
quorum 1
delay 15
}
vrrp_track_process haproxy_portal {
process /usr/sbin/haproxy -Ws -f
/usr/local/pf/var/conf/haproxy-portal.conf -p
/usr/local/pf/var/run/haproxy-portal.pid
full_command
quorum 1
delay 15
}
static_ipaddress {
66.70.255.147 dev lo scope link
}
static_routes {
10.20.254.0/24 via 10.30.247.2 dev eth0.247
10.20.16.0/24 via 10.30.247.2 dev eth0.247
10.20.31.0/24 via 10.30.247.2 dev eth0.247
10.20.253.0/24 via 10.30.247.2 dev eth0.247
10.20.252.0/24 via 10.30.247.2 dev eth0.247
}
*From:* Fabrice Durand <fdur...@inverse.ca>
*Sent:* Friday, October 9, 2020 3:51 PM
*To:* Jeff Linden <jlin...@jerviswebb.com>;
packetfence-users@lists.sourceforge.net
*Subject:* Re: [PacketFence-users] captive_portal.ip_address in
pf.conf.defaults
Can i see the keepalived.conf ?
And do you have something (like error) in the logs about keepalived
(journalctl -f | grep keepalived) when you restart it ?
Le 20-10-09 à 15 h 46, Jeff Linden a écrit :
Keepalived restarts successfully, but is not showing on the lo
interface.
I performed the restart of keepalive using this…
# /usr/local/pf/bin/pfcmd service keepalived restart
Service Status PID
Checking configuration sanity...
packetfence-keepalived.service started 145901
But, no, the address is still not assigned to lo
# ip a
1: lo: <LOOPBACK,UP,LOWER_UP> mtu 65536 qdisc noqueue state
UNKNOWN group default qlen 1
link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
inet 127.0.0.1/8 scope host lo
valid_lft forever preferred_lft forever
inet6 ::1/128 scope host
valid_lft forever preferred_lft forever
Jeff
*From:* Fabrice Durand <fdur...@inverse.ca>
<mailto:fdur...@inverse.ca>
*Sent:* Friday, October 9, 2020 3:30 PM
*To:* Jeff Linden <jlin...@jerviswebb.com>
<mailto:jlin...@jerviswebb.com>;
packetfence-users@lists.sourceforge.net
<mailto:packetfence-users@lists.sourceforge.net>
*Subject:* Re: [PacketFence-users] captive_portal.ip_address in
pf.conf.defaults
When you restart keepalived does the ip appear on lo ?
Does keepalived start ?
Le 20-10-09 à 15 h 20, Jeff Linden a écrit :
Fabrice,
I realized that I tested previously with the line commented
out of pf.conf.defaults.
I’ve put the line back in to pf.conf.defaults and re-run the
test you asked for. Here are the better results. Still no on
the IP being assigned to lo, but yes to it being in the
keepalived.conf.
Does the ip is assigned to lo ? (ip a)
No, it is not assigned to lo. Only 127.0.0.1/8
is assigned
Check the keepalived.conf file if it contain the ip
66.70.255.147 (var/conf/keepalived.conf).
Yes, keepalived.conf does not contain the IP 66.70.255.147
Also check if there is not a keepalived.conf.rpmnew somewhere.
No, there is no keepalived.conf.rpmnew anywhere.
Jeff
*From:* Jeff Linden via PacketFence-users
<packetfence-users@lists.sourceforge.net>
<mailto:packetfence-users@lists.sourceforge.net>
*Sent:* Friday, October 9, 2020 3:10 PM
*To:* Fabrice Durand <fdur...@inverse.ca>
<mailto:fdur...@inverse.ca>;
packetfence-users@lists.sourceforge.net
<mailto:packetfence-users@lists.sourceforge.net>
*Cc:* Jeff Linden <jlin...@jerviswebb.com>
<mailto:jlin...@jerviswebb.com>
*Subject:* Re: [PacketFence-users] captive_portal.ip_address
in pf.conf.defaults
Does the ip is assigned to lo ? (ip a)
No, it is not assigned to lo. Only 127.0.0.1/8
is assigned
Check the keepalived.conf file if it contain the ip
66.70.255.147 (var/conf/keepalived.conf).
No, keepalived.conf does not contain the IP
66.70.255.147
Also check if there is not a keepalived.conf.rpmnew somewhere.
No, there is no keepalived.conf.rpmnew anywhere.
Jeff
*From:* Fabrice Durand <fdur...@inverse.ca
<mailto:fdur...@inverse.ca>>
*Sent:* Friday, October 9, 2020 2:59 PM
*To:* Jeff Linden <jlin...@jerviswebb.com
<mailto:jlin...@jerviswebb.com>>;
packetfence-users@lists.sourceforge.net
<mailto:packetfence-users@lists.sourceforge.net>
*Subject:* Re: [PacketFence-users] captive_portal.ip_address
in pf.conf.defaults
Does the ip is assigned to lo ? (ip a)
Check the keepalived.conf file if it contain the ip
66.70.255.147 (var/conf/keepalived.conf).
Also check if there is not a keepalived.conf.rpmnew somewhere.
Regards
Fabrice
Le 20-10-09 à 14 h 52, Jeff Linden a écrit :
Fabrice,
ps -fe | grep keepalive
root 98543 1 0 13:56 ? 00:00:00
/usr/sbin/keepalived -f
/usr/local/pf/var/conf/keepalived.conf
--pid=/usr/local/pf/var/run/keepalived.pid
root 98549 98543 0 13:56 ? 00:00:00
/usr/sbin/keepalived -f
/usr/local/pf/var/conf/keepalived.conf
--pid=/usr/local/pf/var/run/keepalived.pid
root 98550 98543 0 13:56 ? 00:00:00
/usr/sbin/keepalived -f
/usr/local/pf/var/conf/keepalived.conf
--pid=/usr/local/pf/var/run/keepalived.pid
root 115221 111126 0 14:45 pts/0 00:00:00 grep
keepalive
Keep alive is running fine. I didn’t mention it before,
but I can see those log entries presented below from
haproxy.log are repeating over and over.
And, as I run the systemctl status command I can see the
PID change and the time since it started activating
updates as well.
In the web interface, when I tell the service to stop, it
immediately restarts in the same state I describe below.
Managed, Active, but not Alive.
Additionally, there is a log entry in packetfence.log that
is repeating each time the haproxy-portal service tries to
start. It says “packetfence: -e(82711) WARN: requesting
member ips for an undefined interface...
(pf::cluster::members_ips)”.
Jeff Linden | Corporate Infrastructure Specialist
*DAIFUKU NORTH AMERICA*
30100 Cabot Drive, Novi MI 48377
(248) 553-1234 x1013
*DAIFUKU * <http://www.daifukuna.com/>
*Always an Edge Ahead*
*From:* Fabrice Durand via PacketFence-users
<packetfence-users@lists.sourceforge.net>
<mailto:packetfence-users@lists.sourceforge.net>
*Sent:* Friday, October 9, 2020 2:18 PM
*To:* packetfence-users@lists.sourceforge.net
<mailto:packetfence-users@lists.sourceforge.net>
*Cc:* Fabrice Durand <fdur...@inverse.ca>
<mailto:fdur...@inverse.ca>
*Subject:* Re: [PacketFence-users]
captive_portal.ip_address in pf.conf.defaults
Hello Jeff,
your issue is because keepalived is not running.
let's try:
/usr/local/pf/bin/pfcmd service pf updatesystemd
systemctl restart packetfence-keepalived.service
Regards
Fabrice
Le 20-10-09 à 14 h 11, Jeff Linden via PacketFence-users a
écrit :
Hello,
I’ve upgraded PacketFence from 9.2 to 10.1. Since
then, I’ve had trouble getting the Captive Portal to
function. Since I noticed a newer version is
available, I have now upgraded to 10.2 before writing
this.
In the web interface, under Status -> Services, the
haproxy-portal is enabled and running. All green.
Except, the pid is 0.
Also in the web interface, under Advanced Access
Configuration -> Captive Portal, the haproxy-portal
dropdown is showing green. But, looking further by
clicking the dropdown, I notice Enabled and Managed
are green, but Alive is red.
Systemctl status packetfence-haproxy-portal returns
the following result:
● packetfence-haproxy-portal.service - PacketFence
HAProxy Load Balancer for the captive portal
Loaded: loaded
(/lib/systemd/system/packetfence-haproxy-portal.service;
enabled; vendor preset: enabled)
Active: activating (start-pre) since Fri 2020-10-09
10:57:14 EDT; 2s ago
Process: 230643 ExecStart=/usr/sbin/haproxy -Ws -f
/usr/local/pf/var/conf/haproxy-portal.conf -p
/usr/local/pf/var/run/haproxy-portal.pid (code=exited,
status=1/FAILU
Main PID: 230643 (code=exited, status=1/FAILURE);
Control PID: 230652 (perl)
Tasks: 1 (limit: 36864)
CGroup:
/packetfence.slice/packetfence-haproxy-portal.service
└─control
└─230652 /usr/bin/perl -I/usr/local/pf/lib
-Mpf::services::manager::haproxy_portal -e
pf::services::manager::haproxy_portal->new()->generateConfig()
Oct 09 10:57:16 nadc1-pfence-01 haproxy[230643]:
[ALERT] 282/105714 (230643) : Starting frontend
portal-http-66.70.255.147: cannot bind socket
[66.70.255.147:80]
Oct 09 10:57:16 nadc1-pfence-01 haproxy[230643]:
[ALERT] 282/105714 (230643) : Starting frontend
portal-https-66.70.255.147: cannot bind socket
[66.60.255.147:443]
Oct 09 10:57:14 nadc1-pfence-01 systemd[1]:
packetfence-haproxy-portal.service: Main process
exited, code=exited, status=1/FAILURE
Oct 09 10:57:14 nadc1-pfence-01 systemd[1]: Failed to
start PacketFence HAProxy Load Balancer for the
captive portal.
Oct 09 10:57:14 nadc1-pfence-01 systemd[1]:
packetfence-haproxy-portal.service: Unit entered
failed state.
Oct 09 10:57:14 nadc1-pfence-01 systemd[1]:
packetfence-haproxy-portal.service: Failed with result
'exit-code'.
Oct 09 10:57:14 nadc1-pfence-01 systemd[1]:
packetfence-haproxy-portal.service: Service hold-off
time over, scheduling restart.
Oct 09 10:57:14 nadc1-pfence-01 systemd[1]: Stopped
PacketFence HAProxy Load Balancer for the captive portal.
Oct 09 10:57:14 nadc1-pfence-01 systemd[1]: Starting
PacketFence HAProxy Load Balancer for the captive
portal...
In /var/log/haproxy.log, I find the following:
Oct 9 11:48:38 nadc1-pfence-01 haproxy[17789]: Proxy
proxy started.
Oct 9 11:48:38 nadc1-pfence-01 haproxy[17789]: Proxy
static started.
Oct 9 11:48:38 nadc1-pfence-01 haproxy[17789]:
[ALERT] 282/114838 (17789) : Starting frontend
portal-http-66.70.255.147: cannot bind socket
[66.70.255.147:80]
Oct 9 11:48:38 nadc1-pfence-01 haproxy[17789]:
[ALERT] 282/114838 (17789) : Starting frontend
portal-https-66.70.255.147: cannot bind socket
[66.70.255.147:443]
Oct 9 11:48:38 nadc1-pfence-01 haproxy[17789]: Proxy
portal-http-10.30.247.1 started.
Oct 9 11:48:38 nadc1-pfence-01 haproxy[17789]: Proxy
portal-https-10.30.247.1 started.
Oct 9 11:48:38 nadc1-pfence-01 haproxy[17789]: Proxy
10.30.247.1-backend started.
Oct 9 11:48:38 nadc1-pfence-01 haproxy[17789]: Proxy
portal-http-10.30.3.162 started.
Oct 9 11:48:38 nadc1-pfence-01 haproxy[17789]: Proxy
portal-https-10.30.3.162 started.
Oct 9 11:48:38 nadc1-pfence-01 haproxy[17789]: Proxy
10.30.3.162-backend started.
Oct 9 11:48:38 nadc1-pfence-01 haproxy[17789]: Proxy
portal-http-10.30.248.1 started.
Oct 9 11:48:38 nadc1-pfence-01 haproxy[17789]: Proxy
portal-https-10.30.248.1 started.
Oct 9 11:48:38 nadc1-pfence-01 haproxy[17789]: Proxy
10.30.248.1-backend started.
I notice the error about binding to 66.70.255.147.
That is not an IP I recognize, it is certainly not
assigned to any of the interfaces on my system.
I find the address 66.70.255.147 in the
pf.conf.defaults file with the header
# The IP address the portal uses in the registration
and isolation networks.
# This IP address should point to an IP outside the
registration and isolation networks.
# Do not change unless you know what you are doing.
ip_address=66.70.255.147
I found a github entry that discusses the captive
portal IP here
https://github.com/inverse-inc/packetfence/pull/5682
. It says the previous hardcoded address of 192.0.2.1
is removed and an Inverse owned IP is put in its
place. I see that 66.70.255.147 is owned by Ovh
Hosting in Montreal, not Inverse specifically, but I
believe this github entry is talking about the captive
portal section of pf.conf.defaults.
So, I set the address in the Captive Portal web page
to 192.0.2.1 and experience the same results. No
captive portal and the error with the haproxy-portal
service still exists.
Systemctl status packetfence-haproxy-portal now
returns the following result:
● packetfence-haproxy-portal.service - PacketFence
HAProxy Load Balancer for the captive portal
Loaded: loaded
(/lib/systemd/system/packetfence-haproxy-portal.service;
enabled; vendor preset: enabled)
Active: activating (start-pre) since Fri 2020-10-09
10:57:14 EDT; 2s ago
Process: 230643 ExecStart=/usr/sbin/haproxy -Ws -f
/usr/local/pf/var/conf/haproxy-portal.conf -p
/usr/local/pf/var/run/haproxy-portal.pid (code=exited,
status=1/FAILU
Main PID: 230643 (code=exited, status=1/FAILURE);
Control PID: 230652 (perl)
Tasks: 1 (limit: 36864)
CGroup:
/packetfence.slice/packetfence-haproxy-portal.service
└─control
└─230652 /usr/bin/perl -I/usr/local/pf/lib
-Mpf::services::manager::haproxy_portal -e
pf::services::manager::haproxy_portal->new()->generateConfig()
Oct 09 10:57:16 nadc1-pfence-01 haproxy[230643]:
[ALERT] 282/105714 (230643) : Starting frontend
portal-http-192.0.2.1: cannot bind socket [192.0.2.1:80]
Oct 09 10:57:16 nadc1-pfence-01 haproxy[230643]:
[ALERT] 282/105714 (230643) : Starting frontend
portal-https-192.0.2.1: cannot bind socket [192.0.2.1:443]
Oct 09 10:57:14 nadc1-pfence-01 systemd[1]:
packetfence-haproxy-portal.service: Main process
exited, code=exited, status=1/FAILURE
Oct 09 10:57:14 nadc1-pfence-01 systemd[1]: Failed to
start PacketFence HAProxy Load Balancer for the
captive portal.
Oct 09 10:57:14 nadc1-pfence-01 systemd[1]:
packetfence-haproxy-portal.service: Unit entered
failed state.
Oct 09 10:57:14 nadc1-pfence-01 systemd[1]:
packetfence-haproxy-portal.service: Failed with result
'exit-code'.
Oct 09 10:57:14 nadc1-pfence-01 systemd[1]:
packetfence-haproxy-portal.service: Service hold-off
time over, scheduling restart.
Oct 09 10:57:14 nadc1-pfence-01 systemd[1]: Stopped
PacketFence HAProxy Load Balancer for the captive portal.
Oct 09 10:57:14 nadc1-pfence-01 systemd[1]: Starting
PacketFence HAProxy Load Balancer for the captive
portal...
/var/log/haproxy.log now shows:
Oct 9 10:47:56 nadc1-pfence-01 haproxy[223396]: Proxy
proxy started.
Oct 9 10:47:56 nadc1-pfence-01 haproxy[223396]:
[ALERT] 282/104756 (223396) : Starting frontend
portal-http-192.0.2.1: cannot bind socket [192.0.2.1:80]
Oct 9 10:47:56 nadc1-pfence-01 haproxy[223396]:
[ALERT] 282/104756 (223396) : Starting frontend
portal-https-192.0.2.1: cannot bind socket [192.0.2.1:443]
Oct 9 10:47:56 nadc1-pfence-01 haproxy[223396]: Proxy
static started.
Oct 9 10:47:56 nadc1-pfence-01 haproxy[223396]: Proxy
portal-http-10.30.247.1 started.
Oct 9 10:47:56 nadc1-pfence-01 haproxy[223396]: Proxy
portal-https-10.30.247.1 started.
Oct 9 10:47:56 nadc1-pfence-01 haproxy[223396]: Proxy
10.30.247.1-backend started.
Oct 9 10:47:56 nadc1-pfence-01 haproxy[223396]: Proxy
portal-http-10.30.3.162 started.
Oct 9 10:47:56 nadc1-pfence-01 haproxy[223396]: Proxy
portal-https-10.30.3.162 started.
Oct 9 10:47:56 nadc1-pfence-01 haproxy[223396]: Proxy
10.30.3.162-backend started.
Oct 9 10:47:56 nadc1-pfence-01 haproxy[223396]: Proxy
portal-http-10.30.248.1 started.
Oct 9 10:47:56 nadc1-pfence-01 haproxy[223396]: Proxy
portal-https-10.30.248.1 started.
Oct 9 10:47:56 nadc1-pfence-01 haproxy[223396]: Proxy
10.30.248.1-backend started.
In the pf.conf.defaults file, I commented out the IP.
This produces a warning when restarting the services
“pf.conf value captive_portal.ip_address is not defined!”.
The haproxy-portal service is now started and I
successfully performed guest registration.
Sorry to trouble you with all of this, but the first
time I performed these steps, I was still experiencing
trouble with the captive portal. It’s not until I
went through it all again to collect the information
to include with my question that I found the captive
portal to be working. It is working with the
captive_portal.ip_address section of pf.conf.defaults
commented out. I’m not certain commenting this line
is the correct solution. It must be there for a
reason, no?
I will leave these questions for the group then…
Why is the haproxy-portal showing green in the web
interface when, in fact, it is not successfully started?
What is the story with the captive_portal.ip_address
section of pf.conf.defaults? Is it a mistake to leave
it commented?
Thank you,
Jeff
PRIVACY NOTICE: The information contained in this
e-mail, including any attachments, is confidential and
intended only for the named recipient(s). Unauthorized
use, disclosure, forwarding, or copying is strictly
prohibited and may be unlawful. If you are not the
intended recipient, please delete the e-mail and any
attachments and notify us immediately by return e-mail.
_______________________________________________
PacketFence-users mailing list
PacketFence-users@lists.sourceforge.net
<mailto:PacketFence-users@lists.sourceforge.net>
https://lists.sourceforge.net/lists/listinfo/packetfence-users
--
Fabrice Durand
fdur...@inverse.ca <mailto:fdur...@inverse.ca> :: +1.514.447.4918
(x135) ::www.inverse.ca <http://www.inverse.ca>
Inverse inc. :: Leaders behind SOGo (http://www.sogo.nu) and
PacketFence (http://packetfence.org)
--
Fabrice Durand
fdur...@inverse.ca <mailto:fdur...@inverse.ca> :: +1.514.447.4918 (x135)
::www.inverse.ca <http://www.inverse.ca>
Inverse inc. :: Leaders behind SOGo (http://www.sogo.nu) and
PacketFence (http://packetfence.org)
--
Fabrice Durand
fdur...@inverse.ca <mailto:fdur...@inverse.ca> :: +1.514.447.4918 (x135)
::www.inverse.ca <http://www.inverse.ca>
Inverse inc. :: Leaders behind SOGo (http://www.sogo.nu) and PacketFence
(http://packetfence.org)
--
Fabrice Durand
fdur...@inverse.ca <mailto:fdur...@inverse.ca> :: +1.514.447.4918 (x135)
::www.inverse.ca <http://www.inverse.ca>
Inverse inc. :: Leaders behind SOGo (http://www.sogo.nu) and PacketFence
(http://packetfence.org)