Dear HA members,
I am monitoring on my master and slave machine for a while, i found that
from the tcpdump result, both master/slave is sending udp packet.
But suddenly, the slave take over the master become active node. It
doesn't make any sense :-)
Is there a method that I can fix master always as a active node, the
slave only active when the master is fail.
Now both of my machines keep switching between them :-)
Any guidances or ideas are welcome. Thanks again.
Below is the tcpdump and netstat result.
[EMAIL PROTECTED] ~]# tcpdump -i eth1
tcpdump: verbose output suppressed, use -v or -vv for full protocol decode
listening on eth1, link-type EN10MB (Ethernet), capture size 96 bytes
19:18:44.304269 IP master,nav6.org.53374 > 192.168.1.255.ha-cluster:
UDP, length 185
19:18:45.295768 IP slave.nav6.org.42873 > 192.168.1.255.ha-cluster: UDP,
length 184
19:18:46.303541 IP master,nav6.org.53374 > 192.168.1.255.ha-cluster:
UDP, length 185
19:18:47.294960 IP slave.nav6.org.42873 > 192.168.1.255.ha-cluster: UDP,
length 184
19:18:48.302871 IP master,nav6.org.53374 > 192.168.1.255.ha-cluster:
UDP, length 185
19:18:49.294154 IP slave.nav6.org.42873 > 192.168.1.255.ha-cluster: UDP,
length 184
19:18:50.302207 IP master,nav6.org.53374 > 192.168.1.255.ha-cluster:
UDP, length 185
[EMAIL PROTECTED] ~]# netstat -anp|more
Active Internet connections (servers and established)
Proto Recv-Q Send-Q Local Address Foreign
Address S
tate PID/Program name
tcp 0 0 0.0.0.0:111
0.0.0.0:* L
ISTEN 2184/portmap
tcp 0 0 10.207.160.43:80
0.0.0.0:* L
ISTEN 3142/httpd
tcp 0 0 127.0.0.1:631
0.0.0.0:* L
ISTEN 2469/cupsd
tcp 0 0 0.0.0.0:695
0.0.0.0:* L
ISTEN 2209/rpc.statd
tcp 0 0 127.0.0.1:25
0.0.0.0:* L
ISTEN 2607/sendmail: acce
tcp 0 0 10.207.160.43:80
10.207.161.212:1768 T
IME_WAIT -
tcp 0 0 10.207.160.43:80
10.207.161.212:1769 T
IME_WAIT -
tcp 0 0 10.207.160.43:80
10.207.161.212:1767 T
IME_WAIT -
tcp 0 0 10.207.160.43:80
10.207.161.212:1764 T
IME_WAIT -
tcp 0 0 10.207.160.43:80
10.207.161.212:1765 T
--
Best regards,
Simon Teh
Network and System Administrator
National Advanced IPv6
Centre of Excellence,
School of Computer Science,
Universiti Sains Malaysia
[EMAIL PROTECTED] wrote:
Send Linux-HA mailing list submissions to
[email protected]
To subscribe or unsubscribe via the World Wide Web, visit
http://lists.linux-ha.org/mailman/listinfo/linux-ha
or, via email, send a message with subject or body 'help' to
[EMAIL PROTECTED]
You can reach the person managing the list at
[EMAIL PROTECTED]
When replying, please edit your Subject line so it is more specific
than "Re: Contents of Linux-HA digest..."
Today's Topics:
1. Heartbeat over FibreChannel (Todd, Conor)
2. Re: Heartbeat over FibreChannel (Robert)
3. Slave always become active node - problem (chteh)
4. Re: Slave always become active node - problem (Bj?rn Boschman)
5. AW: [Linux-HA] Restart behavior of master/slave resources
(Krauth, Alexander)
----------------------------------------------------------------------
Message: 1
Date: Tue, 19 Aug 2008 21:50:03 +0000
From: "Todd, Conor" <[EMAIL PROTECTED]>
Subject: [Linux-HA] Heartbeat over FibreChannel
To: General Linux-HA mailing list <[email protected]>
Message-ID:
<[EMAIL PROTECTED]>
Content-Type: text/plain; charset="us-ascii"
Hi. Perhaps this is a moronic question, but would it be possible to use a
Fibre Channel interface (and the storage network its connected to) to send
heartbeat signals? I have a services cluster with a bunch of storage attached,
but it's all connected via SAS, and so all of the nodes are connected to an
unused FC switch. I'd love to make use of it...
- Conor
------------------------------
Message: 2
Date: Wed, 20 Aug 2008 09:04:25 +0200
From: Robert <[EMAIL PROTECTED]>
Subject: Re: [Linux-HA] Heartbeat over FibreChannel
To: General Linux-HA mailing list <[email protected]>
Message-ID: <[EMAIL PROTECTED]>
Content-Type: text/plain; charset=ISO-8859-1; format=flowed
You can run IP over Fibre Channel. So I'd expect heartbeats to work with
IPoverFC.
However I have not tried this - so you may have to google a little.
Robert
Todd, Conor schrieb:
Hi. Perhaps this is a moronic question, but would it be possible to use a
Fibre Channel interface (and the storage network its connected to) to send
heartbeat signals? I have a services cluster with a bunch of storage attached,
but it's all connected via SAS, and so all of the nodes are connected to an
unused FC switch. I'd love to make use of it...
------------------------------
Message: 3
Date: Wed, 20 Aug 2008 15:23:15 +0800
From: chteh <[EMAIL PROTECTED]>
Subject: [Linux-HA] Slave always become active node - problem
To: [email protected]
Message-ID: <[EMAIL PROTECTED]>
Content-Type: text/plain; charset=ISO-8859-1; format=flowed
Dear Ha members,
Apologies if this has been asked before.
I'm new in HA configuration, I am configuring HA for two machines (http tset),
one act as master and the other one act as slave.
Each machine has 2 NIC, one is connected to local LAN and the other one is to
connect master/slave using cross cable.
Master- eth0: 10.207.160.40
eth1: 192.168.1.1
Slave - eth0: 10.207.160.41
eth1: 192.168.1.2
But i have a problem for my setup, my slave machine always likely to overtake
the master to become an Active node,
meanwhile the master will turn to standby mode.
I have checked the master, I am 100% sure that the master machine is still alive and connectivity is ok.
My configuration is as below:
ha.cf
------
use_logd on
logfile /var/log/ha-log
debugfile /var/log/ha-debug
keepalive 2
deadtime 100
initdead 120
bcast eth1
udpport 694
auto_failback on
node master
node slave
haresource
------------
master 10.207.160.43 httpd
authkeys
---------
auth 2
2 crc
Am i miss out anything? I have refer to many different How-To, but still
can't get it done the way I want to.
Hope i could some guidance from here and thanks in advanced.
_______________________________________________
Linux-HA mailing list
[email protected]
http://lists.linux-ha.org/mailman/listinfo/linux-ha
See also: http://linux-ha.org/ReportingProblems
