[Openstack-operators] Measuring how fast is my Cloud with OpenStack
Tim.Bell at cern.ch
Tue Nov 24 18:46:04 UTC 2015
How fast can be measured in a variety of ways:
- How quickly a VM can be spawned and become available ?
- How quickly it runs once it is available ? How many X workunits/second can be achieved with N cores ?
Rally will help with the 1st case. For the second case, it is important to choose an appropriate benchmark for your intended workload. Equally, it needs to be viewed statistically... we see significant performance variations in a single cloud from both hardware generations and loaded hypervisors. Defining the best way to measure this is not easy (but I'd be happy to hear from anyone who has cracked it....)
> -----Original Message-----
> From: Munoz, Obed N [mailto:obed.n.munoz at intel.com]
> Sent: 24 November 2015 18:27
> To: openstack-operators at lists.openstack.org
> Subject: [Openstack-operators] Measuring how fast is my Cloud with
> Hi folks,
> I was just wondering if there’s a good way or framework for measuring how
> fast an OpenStack cloud is?
> As far I can see we could divide measurements from 2 perspectives:
> 1. Operators Perspective
> * I think Rally may help on that. https://wiki.openstack.org/wiki/Rally
> * Any other recommendation is more than welcome.
> 2. Customers or Users Perspective
> * I think it refers to the performance of how fast a customer’s application
> can be launched or scaled.
> * Is there an automatic benchmark tool for it?
> I really appreciate if someone of you could share some insights about how
> measuring OpenStack’s provided cloud by taking in consideration these 2
> Obed N Munoz
> Cloud Engineer @ ClearLinux Project
> Open Source Technology Center
> Intel Corporation
> OpenStack-operators mailing list
> OpenStack-operators at lists.openstack.org
-------------- next part --------------
A non-text attachment was scrubbed...
Size: 7349 bytes
Desc: not available
More information about the OpenStack-operators