On 2006-08-31T08:37:55, Andrew Beekhof <[EMAIL PROTECTED]> wrote:
BTW, this is not correct:
> >> <op id="apache_a1_mon" interval="120s" name="monitor"
> >> timeout="60s">
> >> <instance_attributes id="apache_a1_mon_attr">
> >> <attributes>
> >> <nvpair id="apache_a1_mon_attr_0"
> >name="OCF_CHECK_LEVEL"
> >> value="20"/>
> >> </attributes>
> >> </instance_attributes>
> >> </op>
Or, rather, it is correct syntactically for the CIB, but not consistent
with the OCF RA spec. This is going to come through in the environment
as "OCF_RESKEY_OCF_CHECK_LEVEL".
The OCF RA API specifies the monitoring depth should come through as
OCF_CHECK_LEVEL; in the RA meta-data, it is specified as "depth"
attribute on the action - for better or worse, it should then probably
be a "depth" attribute on our op element.
Now, that might not be the best way, and we might wish to just rip this
out of the RA API standard - and instead suggest the "depth" as a
best-current-practice name for this particular instance parameter (for
the monitor op); which would then come through as OCF_RESKEY_depth.
But, the current way is redundant and clearly not correct.
(I learned about this in Alan's tutorial; Alan's tutorial correctly
describes the current code, but the current code is wrong ;-)
Sincerely,
Lars Marowsky-Brée
--
High Availability & Clustering
SUSE Labs, Research and Development
SUSE LINUX Products GmbH - A Novell Business -- Charles Darwin
"Ignorance more frequently begets confidence than does knowledge"
_______________________________________________________
Linux-HA-Dev: [email protected]
http://lists.linux-ha.org/mailman/listinfo/linux-ha-dev
Home Page: http://linux-ha.org/
