What are Trove, Heat, and Ceilometer used for?

asked 2014-09-09 14:16:46 -0600

Robert gravatar image

updated 2014-09-11 13:34:21 -0600

mpetason gravatar image

I am studying about how does OpenStack works, so I can move on with my OpenStack job. So far I have set up some basic deployment of Icehouse and now I would like to make higher level deployment. I have been reading for a while, but I actually do not understand openstack modules such as ceilometer, heat and trove. I would like to get quality answers about:

  1. Why do we need heat, we have similar tools such as nova, keystone, APi and so on?
  2. Why is used trove for, we have MySQL database and Rabbit message server? I have been reading abouttrove on http://docs.openstack.org/admin-guide... but as I mentioned before I do not get it.
  3. I understand a lit of bit ceilometer it is some monitoring and statistic module.

In early OpenStack versions which tools has been replacement for these threes?

edit retag flag offensive close merge delete


And I also do not understand trove components on http://docs.openstack.org/admin-guide... . Also which kind of data store does trove usees?

Robert gravatar imageRobert ( 2014-09-09 14:25:11 -0600 )edit

2 answers

Sort by ยป oldest newest most voted

answered 2014-09-09 15:06:46 -0600

mpetason gravatar image

Heat : Think templates and automation. You have three web servers in front of a database cluster. Your template specifies that there should always be three web servers. Someone else logs into the cloud and deletes one of the web servers. Heat will deploy another web server automatically (if you configured everything correctly.) Heat will also automate deployment of certain stacks if there are templates out there. Such as it will auto deploy a stack for Word Press if you wanted.

Ceilometer : Provides a type of resource monitoring and alarms for configured triggers. If you add this you could use it to generate data about usage for tenants and have data to create a pricing model or charge back. You could also use parts of this with heat for Auto Scaling. You have three web servers, you create a trigger in Ceilometer to let you know when those three servers hit 90% CPU, when that happens trigger an alarm to create another instance. Now you have four web servers instead of three, you could also scale back down when everything drops below 90%.

Trove : Database as a service. Not as familiar with it but it's adding another tool set to help end users configure Databases without having to know Linux Administrator or Server level configurations.

All of these tools are additions to OpenStack to keep it competitive with services such as AWS. You'll notice a large amount of overlap of functions between AWS and OpenStack. It's not always 1:1 and sometimes OpenStack is ahead of AWS with what it is doing, but you'll usually see some type of correlation.

edit flag offensive delete link more


If you do not understand documentation where could I ask these theoretical-documentation questions?

Robert gravatar imageRobert ( 2014-09-11 07:13:51 -0600 )edit

If you want to go deeper then I would suggest joining the IRC room for Trove. It has the developers that can answer questions more in depth. You should get a response depending on how busy they are:


mpetason gravatar imagempetason ( 2014-09-11 13:17:10 -0600 )edit

@mpeatson I tried this IRC it is disccusin and I did not get any info about my problems..

Robert gravatar imageRobert ( 2014-09-17 17:08:15 -0600 )edit

Which room did you go into? You'll have to make sure you get in there when someone is active. It can take a while to get answers sometimes depending on how busy people are.

mpetason gravatar imagempetason ( 2014-09-17 17:17:42 -0600 )edit

I am in these: #openstack #openstack-101 #openstack-glance.

Robert gravatar imageRobert ( 2014-09-17 17:36:37 -0600 )edit

answered 2014-09-13 14:29:09 -0600

annegentle gravatar image

Hi, I'm the documentation technical lead and I'd like to understand how we can improve that page.

Database services are provided by the trove project's code spinning up a VM that has MySQL installed and configured for the user. For a data-intensive application cluster such as Hadoop you'd want the sahara project.

edit flag offensive delete link more


If I undersatnd it well DBaaS is providing database for my instance while running? I hope I get it right. Why would my instance need such a service? Can you explane me a litle bit.

Robert gravatar imageRobert ( 2014-09-17 17:11:33 -0600 )edit

@annegentle Ok, I still don't know what functions does trove have. To whom provides DataBase as a Service? To end user who is running some instances or to end user who has some database driven web page by MySQL and let's say he want to migrate base for one day to OpenStack and then migrate it back?

Robert gravatar imageRobert ( 2014-09-17 17:16:13 -0600 )edit

So I don not like simple answers and so doesn't a lot of people like me, who doesn't understand all these services, APIs and so on. And when I am reading manual there should be a lot of explanations nut just tables, configuration properties and so on.

Robert gravatar imageRobert ( 2014-09-17 17:16:28 -0600 )edit

Get to know Ask OpenStack

Resources for moderators

Question Tools

1 follower


Asked: 2014-09-09 14:16:46 -0600

Seen: 354 times

Last updated: Sep 11 '14