[prev in list] [next in list] [prev in thread] [next in thread] 

List:       linux-ha
Subject:    Re: Newbie questions regarding heartbeat
From:       Alan Robertson <alanr () bell-labs ! com>
Date:       1999-06-12 0:32:37
[Download RAW message or body]

Hi Herbert!


Herbert Lepper wrote:
> 
> Hi,
> 
> I have tried to play around with heartbeat 0.3.1 but the results
> were not as I expected. I assume that I did not configure it properly.
> 
> What I wanted to have is a HA system comprised of two Linux boxes (each
> having only one NIC) each having its own IP-address and there should be
> a third IP-address which is always taken over by the master.
> 
> The two boxes have kernel 2.2.3 and 2.2.9 with SuSE 6.0 both have
> CONFIG_IP_ALIAS=y. Both are connected to the net and through a
> serial line.
> 
> The configuration of linux-1 is as follows:
> 
>  ha.cf:
>    serial  /dev/ttyS0
>    #watchdog /dev/watchdog
>    udp     eth0
>    keepalive 2
>    deadtime 10
>    #hopfudge 1
>    baud    19200
>    #port   1001
> 
>    #
>    node    linux-1
>    node    linux-2
> 
>  ipresources:
>     linux-1 222.22.22.22
>     linux-2 222.22.22.23
> 
> The configuration of linux-2 is as follows:
> 
>  ha.cf:
>    serial  /dev/ttyS1
>    #watchdog /dev/watchdog
>    udp     eth0
>    keepalive 2
>    deadtime 10
>    #hopfudge 1
>    baud    19200
>    #port   1001
> 
>    #
>    node    linux-1
>    node    linux-2
> 
>  ipresources:
>     linux-2 222.22.22.23
>     linux-1 222.22.22.22

The only IP resources you should include should be "moveable" IP
resources.  So, for example, you should say ONLY
	linux-1 222.22.22.24  (If that's your floating address)

The code doesn't care, and doesn't want to know about administrative
addresses (those that don't move around).  What you put in "ipresources"
is the "normal" (both nodes up) configuration of moveable IP addresses. 
So, what I put in, assumes that you want linux-1 to be primary, and
linux-2 to be the backup.

> What is wrong with this configuration, ie where do I have to
> put the third IP address? Also looking at ha-log I constantly
> get the following error:
> 
> HA_ERROR: Invalid fields count bp=[linux-2 1 375fdee8 STATUS unknown
> 0.00]

Could you email me the logs from each machine with the files as
attacments (so they aren't potentially mangled by your mailer).

I haven't seen an error like this.  I'm not sure what's causing it off
hand.  Maybe I released the code with a bug in it :-(   I looked at the
code that prints the "Invalid fields count" message.  It doesn't look
quite like that in the source.  There is no 'bp=' string in the message,
and there should be more like this:  Invalid fields count [%d] in [%s]"
on line 1012 of heartbeat.c  What version of the software did you
install?  Did you install from .tar.gz, i386.rpm, or .src.rpm?  What
distribution/release/architecture of Linux are you running?

My copy doesn't seem to do that.  I'd been testing mostly the ppp-udp
code, but it looks like my two machines still work, even when I
reconfigure them as plain serial links.

I suspect that the rest of the list members would rather we carried this
discussion on in private email.

	-- Alan Robertson
	   alanr@bell-labs.com

------------------------------------------------------------------------------
Linux HA Web Site:
  http://www.henge.com/~alanr/ha/
Linux HA HOWTO:
  http://metalab.unc.edu/pub/Linux/ALPHA/linux-ha/High-Availability-HOWTO.html
------------------------------------------------------------------------------

[prev in list] [next in list] [prev in thread] [next in thread]