On 7/9/07, Adrian Chapela <[EMAIL PROTECTED]> wrote:
Andrew Beekhof escribió:
> On 7/9/07, Adrian Chapela <[EMAIL PROTECTED]> wrote:
>>
>> Andrew Beekhof escribió:
>> > On 7/6/07, Adrian Chapela <[EMAIL PROTECTED]> wrote:
>> >>
>> >> This rule is quite simple but I can't understand well.
>> >>
>> >> <rule id="my_resource:connected:rule" score="-INFINITY"
>> >> boolean_op="or">
>> >> <expression id="my_resource:connected:expr:undefined"
>> >> attribute="pingd" operation="not_defined"/>
>> >> <expression id="my_resource:connected:expr:zero"
>> >> attribute="pingd" operation="lte" value="0"/>
>> >> </rule>
>> >>
>> >> The first expression check if the attribute pingd is defined (I
>> >> think...) and the second expression compares the same attribute
>> with 0.
>> >> My problem is, What it's this attribute ?? Who update its value (I
>> >> think pingd, but the executable or the OCF script)??
>> >
>> >
>> > the OCF script is just a wrapper for the executable
>> >
>> > have you read all of http://www.linux-ha.org/v2/faq/pingd ?
>>
>> Yes, but I don't understand how can pingd update CIB to change the Score
>> with this rule ??
>
>
>
> it doesn't change the rule, it changes a node attribute named "pingd"
Yes, It was I want to say...
>
> the rule then looks this value up for each node the rule is applied on
>
> How must my script do to update a similar rule ??
>>
>> I only need a clone and the with a rule change a score to move the
>> IPAddr resource, can you explain me a bit ??
>
>
> that depends, when would the IPaddr need to move?
Now, I have a OCF script that starts, stops, etc. the resource. This
script evaluate output of the start, status, stop operation and then use
the attrd_updater to change value of an attribute. The rule looks this
attribute and the change the score to move the resource group.
I want to move IPaddr because I have two mysql servers. One is the
master and the other is the slave. Pingd and MySQL must be started in
Are you using a master_slave resource for MySQL?
two nodes at the same time and only the VIP can be in one node (master
node). When master is down, it has a network failure or a resource
failure the VIP must be assigned to the slave. It will be a new master
at this time. At this moment I must execute a script to block the
connections on the mysql port, look for replication state, etc. but
this is another history!
Since 2.1.0^ you can do this with simple colocation rules.
You do this by having two colocation rules:
one with from=VIP to=MySQL to_role=master score=10000, and
one with from=VIP to=MySQL to_role=slave score=100
I made the scores up, but hopefully you get the point
^ Packages from: http://software.opensuse.org/download/server:/ha-clustering
_______________________________________________
Linux-HA mailing list
[email protected]
http://lists.linux-ha.org/mailman/listinfo/linux-ha
See also: http://linux-ha.org/ReportingProblems
