<html>
<head>
<meta content="text/html; charset=utf-8" http-equiv="Content-Type">
</head>
<body bgcolor="#FFFFFF" text="#000000">
<br>
<br>
<div class="moz-cite-prefix">On 2015/11/24 16:41, Qiming Teng wrote:<br>
</div>
<blockquote cite="mid:20151124084116.GA9016@qiming-ThinkCentre-M58p"
type="cite">
<blockquote type="cite">
<pre wrap="">After debugging, I found that former result is not overridden by
another policy.
<a class="moz-txt-link-freetext" href="http://git.openstack.org/cgit/openstack/senlin/tree/senlin/engine/actions/base.py#n441">http://git.openstack.org/cgit/openstack/senlin/tree/senlin/engine/actions/base.py#n441</a>
</pre>
</blockquote>
<pre wrap="">
I think you are not referring to a correct location in the source code.
That line was only about checking whether there are some policies failed
checking.</pre>
</blockquote>
<br>
If a cluster attached two scaling_in policies(policy1 with priority
20 and policy2 with priority 40), <br>
when policy_check function is called, it will do policy1 checking
first and then check policy2.<br>
If any policy failed, it will return with CHECK_ERROR. Is this
conform to your original design?<br>
<br>
<blockquote cite="mid:20151124084116.GA9016@qiming-ThinkCentre-M58p"
type="cite">
<pre wrap="">
</pre>
<blockquote type="cite">
<blockquote type="cite">
<pre wrap="">
2. if a cluster attached a scaling policy with event =
CLUSTER_SCALE_IN, when aCLUSTER_SCALING_OUT action is
triggered, the policy also will be checked, is this reasonable?
/ When a ScalingPolicy is defined, you can use 'event'
property to specify the action type you want the policy to take
effect on, like:
<a class="moz-txt-link-freetext" href="http://git.openstack.org/cgit/openstack/senlin/tree/examples/policies/scaling_policy.yaml#n5">http://git.openstack.org/cgit/openstack/senlin/tree/examples/policies/scaling_policy.yaml#n5</a>
Although a ScalingPolicy will be checked for both
CLUSTER_SCALE_IN and CLUSTER_SCALE_OUT actions, the check routine
will return immediately if the action type is not what it is
expecting.
<a class="moz-txt-link-freetext" href="http://git.openstack.org/cgit/openstack/senlin/tree/senlin/policies/scaling_policy.py#n133/">http://git.openstack.org/cgit/openstack/senlin/tree/senlin/policies/scaling_policy.py#n133/</a>
</pre>
</blockquote>
<pre wrap="">
Yes it's not checked in pre_op, but all ScalingPolicies still will
be checked whether in cooldown.
<a class="moz-txt-link-freetext" href="http://git.openstack.org/cgit/openstack/senlin/tree/senlin/engine/actions/base.py#n431*">http://git.openstack.org/cgit/openstack/senlin/tree/senlin/engine/actions/base.py#n431*</a>
</pre>
</blockquote>
<pre wrap="">
Each policy instances creates its own policy-binding on a cluster. The
cooldown is recorded and then checked there. I can sense something is
wrong, but so far I'm not quite sure I understand the specific use case
that the current logic fails to support.
</pre>
</blockquote>
Fo following case:<br>
A cluster is attached with two policies as follow.<br>
policy1: type=senlin.policy.scaling, cooldown=60s, event:
CLUSTER_SCALE_IN<br>
policy2: type=senlin.policy.scaling, cooldown=300s, event: <font
color="#ff0000">CLUSTER_SCALE_OUT</font><br>
<br>
Then I do following actions<br>
1) senlin cluster-scale-in -c 1 mycluster
---------------------- scale-in ok<br>
2) after 70s, senlin cluster-scale-in -c 1 mycluster
----------------------- scale-in failed, because of policy2 is
still in cooldown.<br>
<br>
Now 'cooldown' is a common property for any kind of policy, I think
this property maybe not necessary for all kind of policy like
LB_POLICY.<br>
<br>
<blockquote cite="mid:20151124084116.GA9016@qiming-ThinkCentre-M58p"
type="cite">
<pre wrap="">
</pre>
<blockquote type="cite">
<blockquote type="cite">
<pre wrap=""> Your suggestion has a problem when I want different cooldown
for each ceilometer/aodh alarms, for following cases, how
should I do?
1. 15% < cpu_util < 30%, scaling_in 1 instance with 300s
cooldown time
2. cpu_util < 15%, scaling_in 2 instances with 600s
cooldown time
For a senlin webhook, could we assign a policy which will be
checked ?
/ User is not allowed to specify the policy when defining a
webhook. The webhook target is decided by target object(cluster or
node) and target action type./
</pre>
</blockquote>
<pre wrap="">
Yes we can define cooldown for each policy, but my meaning is
that each cluster_scaling_in action is only checked by specified
scaling_policy like OS::Heat::ScalingPolicy in heat.
</pre>
</blockquote>
<pre wrap="">
This is a misconcept because a Senlin policy is not a Heat
ScalingPolicy. A Senlin policy is checked before and/or after a
specified action is performed.</pre>
</blockquote>
<br>
I got they are different, I want to know how combine these
operations(e.g. webhook, <br>
ScalePolicy, cluster, ceilometer alarms) to realize the autoscaling
functions like in Heat? <br>
Hu yanyan has given an combinatorial method, but I think this
method doesn't resolve the case.<br>
<blockquote cite="mid:20151124084116.GA9016@qiming-ThinkCentre-M58p"
type="cite">
<pre wrap="">
</pre>
<blockquote type="cite">
<pre wrap=""> 1） In heat, we could define two scaling_in actions(via define
two OS::Heat::ScalingPolicy polices ), each scaling_in action is
checked by one OS::Heat::ScalingPolicy, so each scaling_in action's
cooldown is only checked in one OS::Heat::ScalingPolicy.
2）But in senlin, each scaling_in action will be checked by all
attached scaling_policies, so all scaling_polices' cooldown will be
checked.How does senlin support different cooldown time for each
scaling_in action?
</pre>
</blockquote>
<pre wrap="">
The built-in policy will check if the action is a 'CLUSTER_SCALE_IN'
action or not. The policy is supposed to help you decide the number of
nodes you want to remove from a cluster. If you have a specific
requirement where you want 2 nodes to be removed under one condition, 1
node to be removed under another condition, you will create two webhooks
to explicitly specify that. A scaling policy will respect the 'count'
parameter you speicifed from the webhook.</pre>
</blockquote>
I really want to discuss the cooldown checking for multiple polices
of same type. <br>
Following is the different for autoscaling in Senlin and Heat.<br>
In Senlin:<br>
policy1: type=senlin.policy.scaling, cooldown=60s, event:
CLUSTER_SCALE_IN<br>
policy2: type=senlin.policy.scaling, cooldown=300s, event:
CLUSTER_SCALE_IN<br>
webhook1: count=1, action=CLUSTER_SCALE_IN<br>
webhook2: count=2, action=CLUSTER_SCALE_IN<br>
<br>
trigger webhook1, all policy1's cooldown and policy2's cooldown
will be checked.<br>
trigger webhook2, all policy1's cooldown and policy2's coolddow
will be checked.<br>
<br>
<br>
In Heat<br>
policy1: type=OS::Heat::ScalingPolicy, cooldown=60s,
scaling_adjustment=1 <br>
policy2: type=OS::Heat::ScalingPolicy, cooldown=300s,
scaling_adjustment=2 <br>
policy1 will return a webhook as webhook1.<br>
policy2 will return a webhook as webhook2.<br>
<br>
trigger webhook1, only policy1's cooldown will be checked.<br>
trigger webhook1, only policy1's cooldown will be checked.<br>
<br>
<blockquote cite="mid:20151124084116.GA9016@qiming-ThinkCentre-M58p"
type="cite">
<pre wrap="">
Each policy has its default cooldown value when created. However, you
can easily override that default value when attaching such a policy to a
cluster. See 'senlin help cluster-policy-attach' for more info.
Regards,
Qiming
__________________________________________________________________________
OpenStack Development Mailing List (not for usage questions)
Unsubscribe: <a class="moz-txt-link-abbreviated" href="mailto:OpenStack-dev-request@lists.openstack.org?subject:unsubscribe">OpenStack-dev-request@lists.openstack.org?subject:unsubscribe</a>
<a class="moz-txt-link-freetext" href="http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack-dev">http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack-dev</a>
</pre>
</blockquote>
<br>
<pre class="moz-signature" cols="72">--
Best regards,
Jun Xu</pre>
</body>
</html>