<html>
<head>
<meta http-equiv="content-type" content="text/html; charset=UTF-8">
</head>
<body text="#000000" bgcolor="#FFFFFF">
<p>Prior to Stein, Ceilometer issued a metric named <i>cpu_util</i>,
which I could use to trigger alarms and autoscaling when CPU
utilization was too high.</p>
<p>cpu_util doesn't exist anymore. Instead, we are asked to use
Gnocchi's <i>rate</i> feature. However, when using rates, alarms
on a group of resources require more parameters than just one
metric: Both an aggregation and a reaggregation method are needed.
<br>
</p>
<p>For example, a group of instances that implement "myapp":<tt><br>
</tt></p>
<p><tt>gnocchi measures aggregation -m cpu --reaggregation mean
--aggregation rate:mean --query server_group=myapp
--resource-type instance</tt></p>
<p><tt>A</tt>ctually, this command uses a deprecated API (but from
what I can see, Aodh still uses it). The new way is like this:</p>
<p><tt>gnocchi aggregates --resource-type instance '(aggregate
rate:mean (metric cpu mean))' server_group=myapp</tt></p>
<p>If rate:mean is in the archive policy, it also works the other
way around:<br>
</p>
<p><tt>gnocchi aggregates --resource-type instance '(aggregate mean
(metric cpu rate:mean))' server_group=myapp</tt></p>
<p>Without reaggregation, I get quite unexpected numbers, including
negative CPU rates. If you want to understand why, see this
discussion with one of the Gnocchi maintainers [1].<br>
</p>
<p><b>My problem</b>: Aodh allows me to set an aggregation method,
but not a reaggregation method. How can I create alarms based on
rates? The problem extends to Heat and autoscaling.<br>
</p>
<p>Thanks much,</p>
<p>Bernd.<br>
</p>
<p>[1] <a class="moz-txt-link-freetext" href="https://github.com/gnocchixyz/gnocchi/issues/1044">https://github.com/gnocchixyz/gnocchi/issues/1044</a><br>
</p>
</body>
</html>