Mailing-List: contact users-help@cloudstack.apache.org; run by ezmlm
Precedence: bulk
Reply-To: users@cloudstack.apache.org
Subject: Re: CloudStack Max Number of Hypervisors
To: users@cloudstack.apache.org
References: <CAMPxEH9wEgxozvN3hs=mQS8Hb5r6UVYEBeitah8DwVPUg7w6fw@mail.gmail.com>
 <BBDE9769-2F06-42D9-9E17-79A0E1F9EE54@shapeblue.com>
From: ilya <ilya.mailing.lists@gmail.com>
Message-ID: <170c3965-ee09-a10b-bdcc-8ab8088c3b97@gmail.com>
Date: Fri, 28 Oct 2016 14:08:39 -0700
User-Agent: Mozilla/5.0 (Macintosh; Intel Mac OS X 10.11; rv:45.0)
 Gecko/20100101 Thunderbird/45.4.0
MIME-Version: 1.0
In-Reply-To: <BBDE9769-2F06-42D9-9E17-79A0E1F9EE54@shapeblue.com>
Content-Type: text/plain; charset=utf-8
Content-Transfer-Encoding: 8bit
archived-at: Fri, 28 Oct 2016 21:08:46 -0000

There is no cloudstack zone limit i'm aware off.

The zone itself has no limit.

Your limits will be defined by back-end storage and network topology.

If you offer Physical VLANs segregation, then you have to be mindful of
Spanning Tree issues, in that case in my experience, you may want to
limit a zone to 200 nodes and less than 300 VLANs.

If you are running SDN, then this limit no longer applies and you can
span with ease beyond 200 nodes per zone.

As for CloudStack ability to manage many nodes, there are known
implementations with close to 1000 hypervisors on 2 pair cloudstack
setup - and its not the final limit.

With that said, you will have to tweak DB, Java and CloudStack, as
default configs - wont scale.

Perhaps explain what you are trying to accomplish and how your
environment is laid out?


On 10/26/16 12:22 AM, Dag Sonstebo wrote:
> Hi Tyler,
> 
> Ultimately the number of hypervisors a management server can manage depends on the workload it handles – i.e. number of users, number of VMs, how complex configurations your users use etc. As a rule of thumb you should always have two management servers for redundancy – but we do see customers run hundreds of hypervisors with this setup. On top of this keep in mind the management server is stateless – in other words it is easy to just build additional servers once your load increases – given that your load balancer configuration is healthy.
> 
> The same redundancy point goes for your backed MySQL services – you should always run at least a slave + master setup.
> 
> Hope this helps,
> Dag Sonstebo
> Cloud Architect
> ShapeBlue
> 
> On 26/10/2016, 04:52, "Tyler Wilson" <kupo@linuxdigital.net> wrote:
> 
>     Hello All,
>     
>     Are there any known issues with scaling issues with the controller
>     services? How many hypervisors should I be able to deploy to a zone before
>     running into issues and having to start a new zone?
>     
>     Thanks for any info!
>     
> 
> 
> Dag.Sonstebo@shapeblue.com 
> www.shapeblue.com
> 53 Chandos Place, Covent Garden, London  WC2N 4HSUK
> @shapeblue
>   
>  
>