I am doing a similar thing, where I have a process which spawns up a fixed
set of tasks, each assigned to a specific period of work, so each task
name is constant, and if a task fails another is restarted with the same
name. So I think this approach should work fine.
On 7/10/15, 1:06 PM, Sargun
In all truth, I think the reason this idea even has legs is that the
master/slave labels are not really accurate for the two mesos servers
based on their actual roles in a mesos cluster.
That said, I¹m not certain the roles can each be easily distilled down
into a single catch-all word.
There
This does sound promising. Though if it restricts one to hosts using systemd
than I don¹t see much value in it over Docker. However no need to preclude
it on choice of init process alone, plus it exercises the ContainerInfo
structure some more. +1
From: Tim Chen t...@mesosphere.io
Reply-To:
If you are OK with a fixed number of slaves. You could set the number of
instances equal to the number of slaves, and then set the constraints of the
job to: constraints: [[ hostname, UNIQUE ]].
-Chris
From: Leigh Martell le...@immun.io
Reply-To: user@mesos.apache.org user@mesos.apache.org
4 matches
Mail list logo
