Thanks Craig, I'll have a look into this approach.

It does feel a little flaky though, I suspect it may be easy for things to get 
out of sync, I'll see how we get on.

Thanks!

Aaron

________________________________
From: craig w [[email protected]]
Sent: 12 March 2015 09:19
To: [email protected]
Subject: Re: Deploying containers to every mesos slave node

Perhaps you could query the Mesos API to see how many slaves there are, then 
use that in the request to Marathon.

On Thu, Mar 12, 2015 at 5:09 AM, Michael Neale 
<[email protected]<mailto:[email protected]>> wrote:
It would be idea if there was as a way to specify to marathon the number of 
instances to be a variable representing the mesos slaves (I don't think 
possible right now).
On Thu, 12 Mar 2015 at 8:07 pm craig w 
<[email protected]<mailto:[email protected]>> wrote:
If you know when the scaling occurs (perhaps there's an API you can query or 
maybe it can notify you), then you can update the configuration for the 
application (deployed using marathon) to change the number of instances (via 
the Marathon REST API).

On Thu, Mar 12, 2015 at 5:03 AM, Aaron Carey 
<[email protected]<mailto:[email protected]>> wrote:
Hi Craig,

I'd looked into that, but I was thinking this may cause issues when our cluster 
auto scales up or down, as instances would no longer equal slaves?

Thanks,
Aaron

________________________________
From: craig w [[email protected]<mailto:[email protected]>]
Sent: 12 March 2015 08:57
To: [email protected]<mailto:[email protected]>
Subject: Re: Deploying containers to every mesos slave node

Aaron,

You could use Marathon (a Mesos framework) to deploy a container to each host 
by using constraints [1] and setting the number of instances of the container 
to equal the number of slaves.

[1] constraints - https://mesosphere.github.io/marathon/docs/constraints.html

On Thu, Mar 12, 2015 at 4:54 AM, Aaron Carey 
<[email protected]<mailto:[email protected]>> wrote:
Hi All,

In setting up our cluster, we require things like consul to be running on all 
of our nodes. I was just wondering if there was any sort of best practice (or a 
scheduler perhaps) that people could share for this sort of thing?

Currently the approach is to use salt to provision each node and add 
consul/mesos slave process and so on to it, but it'd be nice to remove the 
dependency on salt.

Thanks,
Aaron



--

https://github.com/mindscratch
https://www.google.com/+CraigWickesser
https://twitter.com/mind_scratch
https://twitter.com/craig_links



--

https://github.com/mindscratch
https://www.google.com/+CraigWickesser
https://twitter.com/mind_scratch
https://twitter.com/craig_links



--

https://github.com/mindscratch
https://www.google.com/+CraigWickesser
https://twitter.com/mind_scratch
https://twitter.com/craig_links

Reply via email to