[Pacemaker] clone ip definition and location stops myresources...

Gianluca Cecchi gianluca.cecchi at gmail.com
Thu May 20 09:04:54 EDT 2010


On Thu, May 20, 2010 at 2:40 PM, Koch, Sebastian <Sebastian.Koch at netzwerk.de
> wrote:

>  Thanks for your hints. I had the same issue and your tips nearly resolved
> it for me. But i got a question. I setted the default timeout and afterwards
> the pingd resource started to work as expected. I had a IPTABLES Rule
> dropping icmp on one node and recieved:
>
>  [snip]
>
> When i remove the dropping iptables rule with IPTABLES –F the ping works
> again BUT the Migration summary doesn’t change. I understood it like this:
> does the pingd fail the node gets a value like the multiplier (100 in my
> case). I thought: okay the pingd resource works on both nodes again so
> therefore both should have a value of = 100. Am i right or did i understand
> it wrong?
>
>  [snip]
>
> primitive pinggw ocf:pacemaker:pingd \
>
>         params host_list="10.1.1.162" multiplier="100" \
>
>         op monitor interval="10"
>
> [snip]
>
> rsc_defaults $id="rsc-options" \
>
>         resource-stickiness="100"
>
>
>
For me it works as expected (ping scores come back), with these differences,
as Vadym suggested:
- using ping instead of pingd as primitive and giving a 200 for multiplier:
primitive pinggw ocf:pacemaker:ping \
params host_list="192.168.101.1" multiplier="200" timeout="3" \
op monitor interval="10" timeout="60" \
op start interval="0" timeout="60"

- setting default stickiness to 1000
rsc_defaults $id="rsc-options" \
resource-stickiness="1000"

HIH,
Gianluca
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.clusterlabs.org/pipermail/pacemaker/attachments/20100520/d2ad50a4/attachment-0001.html>


More information about the Pacemaker mailing list