Keepalived haven't LVS support

Bug #496932 reported by Geordi LaForge
28
This bug affects 4 people
Affects Status Importance Assigned to Milestone
keepalived (Ubuntu)
Fix Released
Medium
Andres Rodriguez
Declined for Karmic by Mathias Gug

Bug Description

Binary package hint: keepalived

The actual version (1.1.17) of keepalived in Ubuntu 9.10-amd64 have a big problem,
silently the package was build without the LVS support so if you put some
configuration dedicated to balance some servers, the keepalived daemon simply
ignore it without any warning.

This is a big problem because no warning and no information about this
change wasn't present...
This is my configuration:

cat /etc/keepalived/keepalived.conf
global_defs {
   notification_email {
     keepalived-oper
   }
   notification_email_from keepalived-oper
   smtp_server 127.0.0.1
   smtp_connect_timeout 30
   router_id FWLVS1
}

vrrp_instance VI_OUT {
    interface ethOUT
    state MASTER
    virtual_router_id 9
    priority 250
    authentication {
        auth_type AH
        auth_pass keep
    }
    virtual_ipaddress {
        172.22.100.9/24
        172.22.100.12/24
        172.22.100.13/24
    }
    preempt_delay 300
}

virtual_server 172.22.100.12 80 {
    delay_loop 10
    lb_algo wlc
    lb_kind NAT
    persistence_timeout 120
    persistence_granularity 255.255.0.0
    protocol TCP
    quorum 1

  real_server 192.168.69.20 80 {
    weight 1
    TCP_CHECK {
      connect_port 80
      connect_timeout 3
    }
  }

  real_server 192.168.69.21 80 {
    weight 1
    TCP_CHECK {
      connect_port 80
      connect_timeout 3
    }
  }
}

if i run the daemon with the --dump config this is the output:

Dec 15 12:52:33 fwlvs1 Keepalived: Starting VRRP child process, pid=1762
Dec 15 12:52:33 fwlvs1 Keepalived_vrrp: Registering Kernel netlink command channel
Dec 15 12:52:33 fwlvs1 Keepalived_vrrp: Registering gratutious ARP shared channel
Dec 15 12:52:33 fwlvs1 Keepalived_vrrp: Opening file '/etc/keepalived/keepalived.conf'.
Dec 15 12:52:33 fwlvs1 Keepalived_vrrp: Configuration is using : 63250 Bytes
Dec 15 12:52:33 fwlvs1 Keepalived_vrrp: ------< Global definitions >------
Dec 15 12:52:33 fwlvs1 Keepalived_vrrp: Router ID = FWLVS1
Dec 15 12:52:33 fwlvs1 Keepalived_vrrp: Smtp server = 127.0.0.1
Dec 15 12:52:33 fwlvs1 Keepalived_vrrp: Smtp server connection timeout = 30
Dec 15 12:52:33 fwlvs1 Keepalived_vrrp: Email notification from = keepalived-oper
Dec 15 12:52:33 fwlvs1 Keepalived_vrrp: Email notification = keepalived-oper
Dec 15 12:52:33 fwlvs1 Keepalived_vrrp: ------< VRRP Topology >------
Dec 15 12:52:33 fwlvs1 Keepalived_vrrp: VRRP Instance = VI_OUT
Dec 15 12:52:33 fwlvs1 Keepalived_vrrp: Want State = MASTER
Dec 15 12:52:33 fwlvs1 Keepalived_vrrp: Runing on device = ethOUT
Dec 15 12:52:33 fwlvs1 Keepalived_vrrp: Virtual Router ID = 9
Dec 15 12:52:33 fwlvs1 Keepalived_vrrp: Priority = 250
Dec 15 12:52:33 fwlvs1 Keepalived_vrrp: Advert interval = 1sec
Dec 15 12:52:33 fwlvs1 Keepalived_vrrp: Preempt delay = 300 secs
Dec 15 12:52:33 fwlvs1 Keepalived_vrrp: Authentication type = IPSEC_AH
Dec 15 12:52:33 fwlvs1 Keepalived_vrrp: Password = keep
Dec 15 12:52:33 fwlvs1 Keepalived_vrrp: Virtual IP = 3
Dec 15 12:52:33 fwlvs1 Keepalived_vrrp: 172.22.100.9/24 brd 172.22.100.9 dev ethOUT scope global
Dec 15 12:52:33 fwlvs1 Keepalived_vrrp: 172.22.100.12/24 brd 172.22.100.12 dev ethOUT scope global
Dec 15 12:52:33 fwlvs1 Keepalived_vrrp: 172.22.100.13/24 brd 172.22.100.13 dev ethOUT scope global
Dec 15 12:52:34 fwlvs1 Keepalived_vrrp: VRRP_Instance(VI_OUT) Transition to MASTER STATE
Dec 15 12:52:35 fwlvs1 Keepalived_vrrp: VRRP_Instance(VI_OUT) Entering MASTER STATE

As You can see the LVSpart was completly ignored, no warning, no message, nothing!

Revision history for this message
Dave Walker (dogatemycomputer) wrote :

Thanks for reporting this bug and any supporting documentation. Since this bug has enough information provided for a developer to begin work, I'm going to mark it as confirmed and let them handle it from here. Thanks for taking the time to make Ubuntu better!

Changed in keepalived (Ubuntu):
status: New → Confirmed
Thierry Carrez (ttx)
Changed in keepalived (Ubuntu):
importance: Undecided → Medium
Revision history for this message
Geordi LaForge (develop-ipv7) wrote :

Thank You Dave,
i can give you an information, probably this isn't a bug generated in the packaging process for ubuntu,
but some problem between keepalived and the kernel source.
I made some test to build a keepalived from the source tar ball, but whit the default kernel in ubuntu, 2.6.31.
I made some test:

Host:dubhe-V.int.mobile.ipv7.org - 12:44:50_17-12-09 - Path:/mnt/storage/laforge-storage/download/program/keepalived-1.1.19
[root:13]#./configure --with-kernel-dir=/usr/src/kernels/linux-2.6.31.8/
[...]
checking for kernel version... 2.6.31
checking net/ip_vs.h usability... no
checking net/ip_vs.h presence... no
checking for net/ip_vs.h... no
configure: WARNING: keepalived will be built without LVS support.

Host:dubhe-V.int.mobile.ipv7.org - 12:47:12_17-12-09 - Path:/mnt/storage/laforge-storage/download/program/keepalived-1.1.19
[root:15]#./configure --with-kernel-dir=/usr/src/kernels/2.6.27.41-170.2.117.fc10.i686.PAE/
[...]
checking for kernel version... 2.6.27
checking net/ip_vs.h usability... yes
checking net/ip_vs.h presence... yes
checking for net/ip_vs.h... yes
checking for IPVS syncd support... yes

This is a quick analysis of the problem, but permit to close the search of the problem:
1) Something changed in the kernel source put keepalived not able to link the ipvs framework
2) Something changed in the keepalived that break the configure script

Happy to power up Ubuntu ;)

Revision history for this message
Andres Rodriguez (andreserl) wrote :

Hi Geordi.

This is indeed a bug in the configure.in file of the package. However, this bug has already been fixed in Debian version 1.1.17-2 (In karmic we have 1.1.17-1ubuntu1). To get this fixed in Karmic we'll have to backport keepalived 1.1.17-2 into Karmic, from Lucid. However, keepalived 1.1.17-2 is not in Lucid yet. This is because I'm personally waiting for Debian to see if they update to a newer upstream release.

If they don't, I'll personally be requesting the merge and the backport into karmic to fix this bug.

Cheers,

Andres.

Revision history for this message
Geordi LaForge (develop-ipv7) wrote :

Thank You Andreas,
for me isn't a problem because in the meanwhile i'm using the debian package on the servers.
I'm very happy to see that the problem will be resolved soon in the Ubuntu stream.
Thank You very much!

Have a nice day guys!

Revision history for this message
Edwin K (edwin-kalle) wrote :

Meanwhile here's a patch:

diff -Nur keepalived-1.1.17/configure.in keepalived-1.1.17.new/configure.in
--- keepalived-1.1.17/configure.in 2009-02-15 21:12:21.000000000 +0100
+++ keepalived-1.1.17.new/configure.in 2010-01-22 10:56:33.259617929 +0100
@@ -102,7 +102,7 @@
 dnl ----[ Checks for LVS and VRRP support ]----
 IPVS_SUPPORT="_WITHOUT_LVS_"
 if test "$enable_lvs" != "no"; then
- AC_CHECK_HEADER([net/ip_vs.h],
+ AC_CHECK_HEADER([linux/ip_vs.h],
     [IPVS_SUPPORT="_WITH_LVS_"],
     [
       IPVS_SUPPORT="_WITHOUT_LVS_"
diff -Nur keepalived-1.1.17/keepalived/include/check_data.h keepalived-1.1.17.new/keepalived/include/check_data.h
--- keepalived-1.1.17/keepalived/include/check_data.h 2009-03-05 09:27:20.000000000 +0100
+++ keepalived-1.1.17.new/keepalived/include/check_data.h 2010-01-22 10:57:42.931617947 +0100
@@ -41,7 +41,7 @@
 #else
 #define SCHED_MAX_LENGTH IP_VS_SCHEDNAME_MAXLEN
 #endif
-#include <net/ip_vs.h>
+#include <linux/ip_vs.h>
 #else
 #define SCHED_MAX_LENGTH 1
 #endif
diff -Nur keepalived-1.1.17/keepalived/include/ipvswrapper.h keepalived-1.1.17.new/keepalived/include/ipvswrapper.h
--- keepalived-1.1.17/keepalived/include/ipvswrapper.h 2009-03-05 09:28:09.000000000 +0100
+++ keepalived-1.1.17.new/keepalived/include/ipvswrapper.h 2010-01-22 10:58:05.623617909 +0100
@@ -48,7 +48,7 @@
 #elif _KRNL_2_6_
   #include "../libipvs-2.6/libipvs.h"
 #endif
- #include <net/ip_vs.h>
+ #include <linux/ip_vs.h>
 #endif

 #ifndef IP_VS_TEMPLATE_TIMEOUT
diff -Nur keepalived-1.1.17/keepalived/libipvs-2.4/libipvs.h keepalived-1.1.17.new/keepalived/libipvs-2.4/libipvs.h
--- keepalived-1.1.17/keepalived/libipvs-2.4/libipvs.h 2002-07-11 21:28:49.000000000 +0200
+++ keepalived-1.1.17.new/keepalived/libipvs-2.4/libipvs.h 2010-01-22 10:58:57.739617851 +0100
@@ -10,7 +10,7 @@
 #ifndef _LIBIPVS_H
 #define _LIBIPVS_H

-#include <net/ip_vs.h>
+#include <linux/ip_vs.h>

 /*
  * The default IPVS_SVC_PERSISTENT_TIMEOUT is a little larger than average
diff -Nur keepalived-1.1.17/keepalived/libipvs-2.6/libipvs.h keepalived-1.1.17.new/keepalived/libipvs-2.6/libipvs.h
--- keepalived-1.1.17/keepalived/libipvs-2.6/libipvs.h 2005-02-14 00:17:08.000000000 +0100
+++ keepalived-1.1.17.new/keepalived/libipvs-2.6/libipvs.h 2010-01-22 11:01:08.439617837 +0100
@@ -10,7 +10,7 @@
 #ifndef _LIBIPVS_H
 #define _LIBIPVS_H

-#include <net/ip_vs.h>
+#include <linux/ip_vs.h>

 #define MINIMUM_IPVS_VERSION_MAJOR 1
 #define MINIMUM_IPVS_VERSION_MINOR 1

Revision history for this message
Geordi LaForge (develop-ipv7) wrote :

Happy to see that someone remember this bug ;)
Any idea when a new package will be available on the repository?

Revision history for this message
Andres Rodriguez (andreserl) wrote :

Hi guys. I thought this was already fixed in Lucid but it seems it is not. I'll work on it tonight or tomorrow :).

Changed in keepalived (Ubuntu):
assignee: nobody → Andres E. Rodriguez Lazo (andreserl)
status: Confirmed → In Progress
Revision history for this message
Andres Rodriguez (andreserl) wrote :

Hello guys.

Keepalived 1.1.17-2 has been uploaded into Lucid. I'll be requesting a backport to have the fix into karmic :). Enjoy it!

Revision history for this message
Lasse Birnbaum Jensen (lasse-sdu) wrote :

Any estimate on the backport ?

Revision history for this message
Andres Rodriguez (andreserl) wrote :

Lasse,

I've keepalived 1.1.17-2 available in my PPA. Could you please test a keepalived/ipvsadm nat config and see if it works. If it does, please let me know, and could you please post the config (including nat, iptables if used), since I cannot seem to make the nat work.

Thank you.

Add the ppa: sudo add-apt-repository ppa:andreser/ha
And then update keepalived.

Revision history for this message
Andres Rodriguez (andreserl) wrote :

mistake of ppa:

sudo add-apt-repository ppa:andreserl/ha

Revision history for this message
Geordi LaForge (develop-ipv7) wrote :

If you want i can test the package with my configuration too...
Let me know if you think useful a test with a production configuration.

Davide

Revision history for this message
Lasse Birnbaum Jensen (lasse-sdu) wrote :

I works just fine in my test environment. Please rush the backport, i dont like devel repositories in my production enviroment ;)

Revision history for this message
Lasse Birnbaum Jensen (lasse-sdu) wrote :

Sorry forgot configuration

ips has been changed, but should be consistent

For the LVS NAT to work you should not add any rules for the VIP in the PRE og POSTROUTING chains, ipvs does this on its own transparently. The rules add in the test setup is for management of the nodes behind the LVS. The rules in the INPUT chains is for the VIP.

iptables nat rules
# Generated by iptables-save v1.4.4 on Wed Feb 17 17:33:46 2010
*nat
:PREROUTING ACCEPT [1717:278248]
:POSTROUTING ACCEPT [1000:68909]
:OUTPUT ACCEPT [1000:68909]
-A PREROUTING -d 10.x.x.x/32 -i eth0 -m state --state NEW -j DNAT --to-destination 10.z.z.z
-A PREROUTING -d 10.x.x.y/32 -i eth0 -m state --state NEW -j DNAT --to-destination 10.z.z.v
-A POSTROUTING -s 10.z.z.z/32 -o eth0 -j SNAT --to-source 10.x.x.x
-A POSTROUTING -s 10.z.z.v/32 -o eth0 -j SNAT --to-source 10.x.x.y
COMMIT
# Completed on Wed Feb 17 17:33:46 2010
# Generated by iptables-save v1.4.4 on Wed Feb 17 17:33:46 2010
*filter
:INPUT ACCEPT [4479:1328198]
:FORWARD ACCEPT [0:0]
:OUTPUT ACCEPT [4396:360856]
-A INPUT -d 10.x.x.VIP/32 -i eth0 -p tcp -m tcp --dport 25 -j ACCEPT
-A INPUT -d 10.x.x.VIP/32 -i eth0 -p icmp -j ACCEPT
-A INPUT -d 10.x.x.VIP/32 -i eth0 -j DROP
COMMIT
# Completed on Wed Feb 17 17:33:46 2010

remember the ip4_forward

echo "1" > /proc/sys/net/ipv4/ip_forward

nat wont work without it

keepalived.conf

vrrp_instance VIP_MXP {
        state MASTER
        interface eth1
        priority 250
        authentication {
                auth_type PASS
                auth_pass verysecretpassword
        }
        virtual_router_id 200
        virtual_ipaddress {
                10.z.z.BVIP/18 dev eth1 # back default gw
                10.x.x.VIP/20 dev eth0 # front vip
                10.x.x.x/20 dev eth0 # front node1
                10.x.x.y/20 dev eth0 # front node2
        }
        preempt_delay 300
        garp_master_delay 5
}

virtual_server 10.x.x.VIP 25 {
    delay_loop 60
    lb_algo wrr
    lb_kind NAT
    persistence_timeout 3600
    persistence_granularity 255.255.240.0
    protocol TCP
    real_server 10.z.z.z 25 { #node 1
        weight 100
        SMTP_CHECK {
                connect_timeout 10
                retry 3
                delay_before_retry 1
        }
    }
    real_server 10.z.z.v 25 { #node 2
        weight 100
        SMTP_CHECK {
                connect_timeout 10
                retry 3
                delay_before_retry 1
        }
    }
}

Revision history for this message
Lasse Birnbaum Jensen (lasse-sdu) wrote :

Note:
i have 2 interfaces with static address, front (eth0) and back (eth1) of the lvs. I use VIPS (VIP and BVIP) in example config on both sides. This ensures that i with multiple loadbalancers can move different instances to different lvs machines thus not having fixed master/slave but i use master/master with multiple instances on each box.

Revision history for this message
Andres Rodriguez (andreserl) wrote :

Outside Network -- eth1 -- LB -- eth2 -- Real Servers

eth1: 192.168.0.50 (VIP1: 192.168.0.100)
eth2:10.10.10.50 (VIP2:10.10.10.1)
Real1: eth1:10.10.10.100
Real2 eth1 = 10.10.10.110

Ok, my config is the following:

global_defs {
   notification_email {
     keepalived-oper
   }
   notification_email_from keepalived-oper
   smtp_server 127.0.0.1
   smtp_connect_timeout 30
   router_id UBUNTULVS1
}

vrrp_sync_group VG1 {
   group {
      VI_IP1
   }
}

vrrp_instance VI_IP1 {
    state MASTER
    interface eth2
    lvs_sync_daemon_interface eth2
    virtual_router_id 9
    priority 250
    authentication {
        auth_type PASS
        auth_pass password
    }
    virtual_ipaddress {
        192.168.0.100/24 dev eth1
        10.10.10.1/24 dev eth2
    }
    preempt_delay 300
}

virtual_server 192.168.0.100 80 {
    delay_loop 10
    lb_algo rr
    lb_kind NAT
    ! nat_mask 255.255.255.0
    persistence_timeout 120
    persistence_granularity 255.255.255.0
    protocol TCP
    quorum 1

  real_server 10.10.10.100 80 {
    weight 1
    TCP_CHECK {
      connect_port 80
      connect_timeout 3
    }
  }

  real_server 10.10.10.110 80 {
    weight 1
    TCP_CHECK {
      connect_port 80
      connect_timeout 3
    }
  }
}

iptables rules are:

iptables -A INPUT -d 192.168.0.100/32 -i eth1 -p tcp -m tcp --dport 80 -j ACCEPT
iptables -A INPUT -d 192.168.0.100/32 -i eth1 -p icmp -j ACCEPT
iptables -A INPUT -d 192.168.0.100/32 -i eth1 -j DROP

I try to access the VIP through web browser and It just wont "loadbalance". You can see that the http request I did goes directly to InActConn. So, i don;t really know what might be the issue (ip_forward is enabled btw...)

root@server1:~# ipvsadm -L -n
IP Virtual Server version 1.2.1 (size=4096)
Prot LocalAddress:Port Scheduler Flags
  -> RemoteAddress:Port Forward Weight ActiveConn InActConn
TCP 192.168.0.100:80 rr persistent 120 mask 255.255.255.0
  -> 10.10.10.100:80 Masq 1 0 1

Any ideas of what might be going wrong? I'd really appreciate your help.

Revision history for this message
Andres Rodriguez (andreserl) wrote :

and this other iptables :

iptables -t nat -A PREROUTING -d 192.168.0.50/32 -i eth1 -m state --state NEW -j DNAT --to-destination 10.10.10.100
iptables -t nat -A POSTROUTING -s 10.10.10.100/32 -o eth1 -j SNAT --to-source 192.168.0.50

Revision history for this message
Lasse Birnbaum Jensen (lasse-sdu) wrote :

Does your backend/real servers use the 10.10.10.1 as gateway? also you should either use lb_algo wrr or drop the weigths on the realservers. keepalived/ipvs had problems with such configurations earlier - i just dropped functionallity without any warning.

1 backend node missing
Look in your /var/log/syslog to see wether the 10.10.10.110 server in dropped because it doesnt answer on port 80

try the command "ipvsadm -L -n -c" this lists active connections seen from ipvs.

Revision history for this message
Andres Rodriguez (andreserl) wrote :
Download full text (3.8 KiB)

Real servers have 10.10.10.1 as gateway. Changed lb_algo. .110 webserver was down intentionally. IP from where Im trying to access the webserver: 192.168.0.51

Feb 18 10:49:44 karmic1 Keepalived: Starting Keepalived v1.1.17 (02/10,2010)
Feb 18 10:49:44 karmic1 Keepalived_healthcheckers: Initializing ipvs 2.6
Feb 18 10:49:44 karmic1 Keepalived_healthcheckers: Using MII-BMSR NIC polling thread...
Feb 18 10:49:44 karmic1 Keepalived_healthcheckers: Registering Kernel netlink reflector
Feb 18 10:49:44 karmic1 Keepalived_healthcheckers: Registering Kernel netlink command channel
Feb 18 10:49:44 karmic1 Keepalived_healthcheckers: Opening file '/etc/keepalived/keepalived.conf'.
Feb 18 10:49:44 karmic1 Keepalived_healthcheckers: Configuration is using : 12239 Bytes
Feb 18 10:49:44 karmic1 Keepalived_healthcheckers: Activating healtchecker for service [10.10.10.100:80]
Feb 18 10:49:44 karmic1 Keepalived_healthcheckers: Activating healtchecker for service [10.10.10.110:80]
Feb 18 10:49:44 karmic1 Keepalived: Starting Healthcheck child process, pid=2731
Feb 18 10:49:44 karmic1 Keepalived: Starting VRRP child process, pid=2732
Feb 18 10:49:44 karmic1 Keepalived_vrrp: Using MII-BMSR NIC polling thread...
Feb 18 10:49:44 karmic1 Keepalived_vrrp: Registering Kernel netlink reflector
Feb 18 10:49:44 karmic1 Keepalived_vrrp: Registering Kernel netlink command channel
Feb 18 10:49:44 karmic1 Keepalived_vrrp: Registering gratutious ARP shared channel
Feb 18 10:49:44 karmic1 Keepalived_vrrp: Initializing ipvs 2.6
Feb 18 10:49:44 karmic1 Keepalived_vrrp: Opening file '/etc/keepalived/keepalived.conf'.
Feb 18 10:49:44 karmic1 Keepalived_vrrp: Configuration is using : 38787 Bytes
Feb 18 10:49:44 karmic1 kernel: [ 496.793507] IPVS: sync thread started: state = MASTER, mcast_ifn = eth2, syncid = 9
Feb 18 10:49:45 karmic1 Keepalived_vrrp: VRRP_Instance(VI_IP1) Transition to MASTER STATE
Feb 18 10:49:46 karmic1 Keepalived_vrrp: VRRP_Instance(VI_IP1) Entering MASTER STATE
Feb 18 10:49:46 karmic1 avahi-daemon[829]: Registering new address record for 192.168.0.100 on eth1.IPv4.
Feb 18 10:49:46 karmic1 avahi-daemon[829]: Registering new address record for 10.10.10.1 on eth2.IPv4.
Feb 18 10:49:46 karmic1 Keepalived_vrrp: VRRP_Group(VG1) Syncing instances to MASTER state
Feb 18 10:49:46 karmic1 Keepalived_vrrp: Netlink: skipping nl_cmd msg...
Feb 18 10:49:46 karmic1 Keepalived_vrrp: Netlink: skipping nl_cmd msg...
Feb 18 10:49:48 karmic1 Keepalived_healthcheckers: TCP connection to [10.10.10.110:80] failed !!!
Feb 18 10:49:48 karmic1 Keepalived_healthcheckers: Removing service [10.10.10.110:80] from VS [192.168.0.100:80]
Feb 18 10:49:48 karmic1 Keepalived_healthcheckers: SMTP connection ERROR to [127.0.0.1:25].
Feb 18 10:51:04 karmic1 Keepalived_healthcheckers: TCP connection to [10.10.10.110:80] success.
Feb 18 10:51:04 karmic1 Keepalived_healthcheckers: Adding service [10.10.10.110:80] to VS [192.168.0.100:80]
Feb 18 10:51:04 karmic1 Keepalived_healthcheckers: Gained quorum 1+0=1 <= 2 for VS [192.168.0.100:80]
Feb 18 10:51:04 karmic1 Keepalived_healthcheckers: SMTP connection ERROR to [127.0.0.1:25].

root@karmic1:~# ipvsadm -L -n -c
IPVS connection entries
pro expire stat...

Read more...

Revision history for this message
Andres Rodriguez (andreserl) wrote :

Everything working in Karmic, going to request a backport.

Revision history for this message
frell (lee) wrote :

add-apt-repository ppa:andreserl/ha working perfect with 9.10 using DIRECT mode.

Thank you Andres.

Revision history for this message
Martin Pitt (pitti) wrote :

Reportedly fixed in lucid, see bug 530945

Changed in keepalived (Ubuntu):
status: In Progress → Fix Released
tags: added: verification-needed
Revision history for this message
Martin Pitt (pitti) wrote : Please test proposed package

Accepted keepalived into karmic-proposed, the package will build now and be available in a few hours. Please test and give feedback here. See https://wiki.ubuntu.com/Testing/EnableProposed for documentation how to enable and use -proposed. Thank you in advance!

To post a comment you must log in.
This report contains Public information  
Everyone can see this information.

Other bug subscribers

Remote bug watches

Bug watches keep track of this bug in other bug trackers.