--- [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. Frustrating Problems With Active/Passive
> Config (Peregrino)
> 2. Re: pingd removed transient attr from node
> attributes after
> short network outage, and did not recreated it
> (Dejan Muhamedagic)
> 3. Re: Frustrating Problems With Active/Passive
> Config
> (Dejan Muhamedagic)
> 4. Re: RE: Missing gratious ARP (Alan Robertson)
> 5. Failure to start resource makes it impossible
> to fail back
> (Anders Brownworth)
> 6. Can't reload grouped resources (An?bal
> Monsalve Salazar)
> 7. clone behavior when set on-fail to "restart"
> (Junko IKEDA)
>
>
>
----------------------------------------------------------------------
>
> Message: 1
> Date: Mon, 12 Nov 2007 10:49:41 -0800 (PST)
> From: Peregrino <[EMAIL PROTECTED]>
> Subject: [Linux-HA] Frustrating Problems With
> Active/Passive Config
> To: [email protected]
> Message-ID:
> <[EMAIL PROTECTED]>
> Content-Type: text/plain; charset=iso-8859-1
>
> Hello all,
>
> I have an Active/Passive configuration that I have
> been trying to get working that is very frustrating.
> This may have already been answered somewhere, and
> if
> so, just point me to the place.
>
> Overview:
> I have a system (PUBLIC_WEB) with a public viewable
> IP
> address (128.249.65.X) that runs a web service. I
> have
> a VMware Host on another system with a public
> viewable
> address (VMWARE_HOST) (128.249.65.X) that contains a
> virtual system (VIRT_PUBLIC_WEB) with a private ip
> (172.30.10.X).
> I want the virtual system (VIRT_PUBLIC_WEB) with the
> private IP (172.30.10.X) to provide simple failover
> for the public system (PUBLIC_WEB).
> I have this working with another node, but in that
> case both nodes are on the private ip (172.30.10.X),
> so I know it is possible for VMware virtual systems
> to
> act as failover devices.
>
> In all cases the host files are correct, the ha.cf
> files match as do the haresources files. All systems
> can ping one another. I am using heartbeat v2, but
> configured as a v1 (no crm or anything fancy).
>
> Questions:
> First of all, is it possible for a heartbeat node on
> one net to failover and assume a different nets
> address. That seems to be the hangup here.
> If anyone else has done a setup like this I would
> appreciate some assistance.
>
> Second, might it be a routing issue? Each system can
> ping across its net to the other, but does heartbeat
> have its own way of doing things that may not
> include
> using the standard linux net tools?
>
> When the master node (PUBLIC_WEB) is taken down, the
> virtual system (VIRT_PUBLIC_WEB) detects it and
> craps
> out running the resource.d/IPaddr <ip address>
> start.
>
> Any help would be appreciated. Once I have been
> assured that this configuration is possible, I will
> post any config files that are needed to help me fix
> this problem.
>
> Many Thank In Advance.
>
> __________________________________________________
> Do You Yahoo!?
> Tired of spam? Yahoo! Mail has the best spam
> protection around
> http://mail.yahoo.com
>
>
Here is the error log when I shutdown the Active node
on the public IP. Remember, the Active node is on a
public IP, and the Passive node is on a private IP.
=====================
Nov 9 13:55:07 opalVM heartbeat: [2019]: WARN: node
opal.brc.bcm.edu: is dead
Nov 9 13:55:07 opalVM heartbeat: [2019]: info:
Comm_now_up(): updating status to active
Nov 9 13:55:07 opalVM heartbeat: [2019]: info: Local
status now set to: 'active'
Nov 9 13:55:07 opalVM heartbeat: [2019]: WARN: No
STONITH device configured.
Nov 9 13:55:07 opalVM heartbeat: [2019]: WARN: Shared
disks are not protected.
Nov 9 13:55:07 opalVM heartbeat: [2019]: info:
Resources being acquired from opal.brc.bcm.edu.
Nov 9 13:55:07 opalVM heartbeat: [2408]: info: No
local resources
[/usr/local/HA-2.0.8/lib/heartbeat/ResourceManager
listkeys opalvm.cluster.brc.bcm.edu] to acquire.
Nov 9 13:55:07 opalVM heartbeat: [2408]: info:
Writing type [resource] message to FIFO
Nov 9 13:55:07 opalVM heartbeat: [2408]: info: FIFO
message [type resource] written rc=79
Nov 9 13:55:07 opalVM harc[2407]: info: Running
/usr/local/HA-2.0.8/etc/ha.d/rc.d/status status
Nov 9 13:55:07 opalVM heartbeat: [2019]: info:
AnnounceTakeover(local 0, foreign 1, reason
'T_RESOURCES' (0))
Nov 9 13:55:07 opalVM heartbeat: [2019]: info:
AnnounceTakeover(local 1, foreign 1, reason
'T_RESOURCES(us)' (0))
Nov 9 13:55:07 opalVM heartbeat: [2019]: info:
Initial resource acquisition complete
(T_RESOURCES(us))
Nov 9 13:55:07 opalVM heartbeat: [2019]: info: STATE
1 => 3
Nov 9 13:55:07 opalVM heartbeat: [2019]: info:
Exiting req_our_resources process 2408 returned rc 0.
Nov 9 13:55:07 opalVM heartbeat: [2019]: info:
AnnounceTakeover(local 1, foreign 1, reason
'req_our_resources' (1))
Nov 9 13:55:07 opalVM mach_down[2427]: info: Taking
over resource group 128.249.65.204
Nov 9 13:55:07 opalVM ResourceManager[2447]: info:
Acquiring resource group: opal.brc.bcm.edu
128.249.65.204 apache::/etc/httpd/conf/httpd.conf
Nov 9 13:55:07 opalVM IPaddr[2471]: INFO: Resource
is stopped
Nov 9 13:55:07 opalVM ResourceManager[2447]: info:
Running /usr/local/HA-2.0.8/etc/ha.d/resource.d/IPaddr
128.249.65.204 start
Nov 9 13:55:07 opalVM IPaddr[2525]: ERROR:
/usr/local/HA-2.0.8/lib/heartbeat/findif
128.249.65.204 failed [rc=1].
Nov 9 13:55:07 opalVM IPaddr[2516]: ERROR: Generic
error
Nov 9 13:55:08 opalVM ResourceManager[2447]: ERROR:
Return code 1 from
/usr/local/HA-2.0.8/etc/ha.d/resource.d/IPaddr
Nov 9 13:55:08 opalVM ResourceManager[2447]: CRIT:
Giving up resources due to failure of 128.249.65.204
Nov 9 13:55:08 opalVM ResourceManager[2447]: info:
Releasing resource group: opal.brc.bcm.edu
128.249.65.204 apache::/etc/httpd/conf/httpd.conf
Nov 9 13:55:08 opalVM ResourceManager[2447]: info:
Running /usr/local/HA-2.0.8/etc/ha.d/resource.d/apache
/etc/httpd/conf/httpd.conf stop
Nov 9 13:55:08 opalVM apache[2584]: INFO: apache is
not running.
Nov 9 13:55:08 opalVM apache[2581]: INFO: Success
Nov 9 13:55:08 opalVM ResourceManager[2447]: info:
Running /usr/local/HA-2.0.8/etc/ha.d/resource.d/IPaddr
128.249.65.204 stop
Nov 9 13:55:09 opalVM IPaddr[2664]: INFO: Success
Nov 9 13:55:09 opalVM mach_down[2427]: info:
/usr/local/HA-2.0.8/lib/heartbeat/mach_down:
nice_failback: foreign resources acquired
Nov 9 13:55:09 opalVM mach_down[2427]: info:
mach_down takeover complete for node opal.brc.bcm.edu.
Nov 9 13:55:09 opalVM heartbeat: [2019]: info:
AnnounceTakeover(local 1, foreign 1, reason
'T_RESOURCES(us)' (1))
Nov 9 13:55:09 opalVM heartbeat: [2019]: info:
mach_down takeover complete.
Nov 9 13:55:09 opalVM heartbeat: [2019]: info:
AnnounceTakeover(local 1, foreign 1, reason
'mach_down' (1))
Nov 9 13:55:09 opalVM heartbeat: [2019]: info:
Exiting status process 2407 returned rc 0.
Nov 9 13:55:09 opalVM setroubleshoot: SELinux is
preventing /sbin/ifconfig (ifconfig_t) "read write" to
socket:[6686] (initrc_t). For complete SELinux
messages. run sealert -l
2def91cb-1661-4e67-a845-ff44f62aa3eb
Nov 9 13:55:17 opalVM heartbeat: [2019]: info: Local
Resource acquisition completed. (none)
Nov 9 13:55:17 opalVM heartbeat: [2019]: info: local
resource transition completed.
Nov 9 13:55:17 opalVM heartbeat: [2019]: info:
AnnounceTakeover(local 1, foreign 1, reason
'T_RESOURCES(us)' (1))
Nov 9 13:55:39 opalVM hb_standby[2717]: Going standby
[foreign].
Nov 9 13:55:39 opalVM heartbeat: [2019]: info:
opalvm.cluster.brc.bcm.edu wants to go standby
[foreign]
Nov 9 13:55:39 opalVM heartbeat: [2019]: info:
i_hold_resources: 3
Nov 9 13:55:39 opalVM heartbeat: [2019]: info: New
standby state: 1
Nov 9 13:55:40 opalVM heartbeat: [2019]: WARN:
Standby timer has 9610 ms left
===============
Here is my ha.cf file from the passive node on the
private IP (it matches the active nodes ha.cf)
==========================
logfacility daemon
keepalive 1
deadtime 10
warntime 5
initdead 120 # depend on your hardware
udpport 694
ucast eth0 192.168.1.22
auto_failback off
node opal.brc.bcm.edu
node opalVM.cluster.brc.bcm.edu
crm no
/usr/local/HA-2.0.8/lib/heartbeat/ipfail
debug 2
=====================
And here is my haresources file from the passive node
on the private IP.
=====================
opal.brc.bcm.edu 128.249.65.204
apache::/etc/httpd/conf/httpd.conf
=====================
This Active/Passive config works on a test system
where both nodes are on the same private network
(172.30.10.X), but it is not working here with each
node on a separate network (128.249.65.X and
172.30.10.X).
Note the following errors when heartbeat tries to
reassign the IP address:
Nov 9 13:55:07 opalVM ResourceManager[2447]: info:
Running /usr/local/HA-2.0.8/etc/ha.d/resource.d/IPaddr
128.249.65.204 start
Nov 9 13:55:07 opalVM IPaddr[2525]: ERROR:
/usr/local/HA-2.0.8/lib/heartbeat/findif
128.249.65.204 failed [rc=1].
Nov 9 13:55:07 opalVM IPaddr[2516]: ERROR: Generic
error
Nov 9 13:55:08 opalVM ResourceManager[2447]: ERROR:
Return code 1 from
/usr/local/HA-2.0.8/etc/ha.d/resource.d/IPaddr
Nov 9 13:55:08 opalVM ResourceManager[2447]: CRIT:
Giving up resources due to failure of 128.249.65.204
---------------------
I have 3 physical NIC's (eth0, eth1, eth2).
eth0 is the heartbeat detection.
eth1 is on a VLAN for the public IP
eth2 is on a VLAN for the private IP.
Both nodes can ping all other nodes in both networks
(Public=128.249.65.X, Private=172.30.10.X, and
Heartbeat=192.10.1.X)
----------------------
If there is anything else anyone needs to help, please
let me know.
____________________________________________________________________________________
Get easy, one-click access to your favorites.
Make Yahoo! your homepage.
http://www.yahoo.com/r/hs
_______________________________________________
Linux-HA mailing list
[email protected]
http://lists.linux-ha.org/mailman/listinfo/linux-ha
See also: http://linux-ha.org/ReportingProblems
