<div dir="ltr"><div><div>I solved the issue:<br><br>I had to configure resrouce-stickness option to 0 and now whatever hosts I reboot the resources are back and stay like this:<br>     ClusterVIP:0       (ocf::heartbeat:IPaddr2):       Started haproxy01<br>     ClusterVIP:1       (ocf::heartbeat:IPaddr2):       Started haproxy02<br><br></div>Regads<br></div>Thanks for your time and support.<br></div><div class="gmail_extra"><br><div class="gmail_quote">2015-02-06 15:04 GMT-03:00 Michael Schwartzkopff <span dir="ltr"><<a href="mailto:ms@sys4.de" target="_blank">ms@sys4.de</a>></span>:<br><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex">Am Freitag, 6. Februar 2015, 08:14:22 schrieben Sie:<br>
<span class="">> Hi Michael.<br>
><br>
> I mean this, when I reboot the server., pacemaker starts automatically but<br>
> I do no see the VIP in the eth0, sorry I used the word alias which may have<br>
> lead to a confision.<br>
><br>
> ip addr show<br>
> 1: lo: <LOOPBACK,UP,LOWER_UP> mtu 16436 qdisc noqueue state UNKNOWN<br>
>     link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00<br>
>     inet <a href="http://127.0.0.1/8" target="_blank">127.0.0.1/8</a> scope host lo<br>
>     inet6 ::1/128 scope host<br>
>        valid_lft forever preferred_lft forever<br>
><br>
> 2: eth0: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc pfifo_fast state<br>
> UP qlen 1000<br>
>     link/ether 00:21:f6:00:00:25 brd ff:ff:ff:ff:ff:ff<br>
>     inet <a href="http://1.1.1.2/16" target="_blank">1.1.1.2/16</a> brd 1.1.255.255 scope global eth0 ( example server<br>
> service IP )<br>
>     inet <a href="http://1.1.1.1/16" target="_blank">1.1.1.1/16</a> brd 1.1.255.255 scope global secondary eth0 ( THIS IS<br>
> THE VIP) which is not created/configured  automatically.<br>
><br>
> So I have to start it with<br>
</span>> *:/usr/sbin/crm_resource --meta -t primitive -r res_IPaddr2_virtualip -p<br>
> target-role -v started*<br>
<span class="">> Or from the LCMC java frontend I have to perform an stop/start.<br>
><br>
> The output of crm_mon looks like this on both  situations, the resources<br>
> are shown as started but the VIP is not brought up<br>
> automatically when the servers boots up eventhough pacemaker starts<br>
> automatically.<br>
><br>
> Clone Set: cl_IPaddr2_1 [res_IPaddr2_virtualip] (unique)<br>
>      res_IPaddr2_virtualip:0    (ocf::heartbeat:IPaddr2):       Started<br>
> haproxy02<br>
>      res_IPaddr2_virtualip:1    (ocf::heartbeat:IPaddr2):       Started<br>
> haproxy02<br>
<br>
</span>Are you sure that the second IP address res_IPaddr_vitualip:1 no not started?<br>
Perhaps it is only started on the second node? As you see above both instances<br>
of the clone ...:0 and ...:1 are started in the same node haproxy02.<br>
<div class="HOEnZb"><div class="h5"><br>
Mit freundlichen Grüßen,<br>
<br>
Michael Schwartzkopff<br>
<br>
--<br>
[*] sys4 AG<br>
<br>
<a href="http://sys4.de" target="_blank">http://sys4.de</a>, +49 (89) 30 90 46 64, +49 (162) 165 0044<br>
Franziskanerstraße 15, 81669 München<br>
<br>
Sitz der Gesellschaft: München, Amtsgericht München: HRB 199263<br>
Vorstand: Patrick Ben Koetter, Marc Schiffbauer<br>
Aufsichtsratsvorsitzender: Florian Kirstein</div></div></blockquote></div><br></div>