[
https://issues.apache.org/jira/browse/MESOS-5377?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15358190#comment-15358190
]
Benjamin Mahler commented on MESOS-5377:
----------------------------------------
Mitigations have been provided via a GPU framework capability (MESOS-5634)
(which of course, GPU specific) and by allowing operators to exclude resources
from fair sharing (see MESOS-5758).
The GPU framework capability helps to reduce the likelihood that non-GPU
workloads starve out GPU workloads that want to run on the GPU machines. There
are caveats to this, for example:
(1) If the framework is non-cooperative, it may fill GPU machines with non-GPU
workloads, and there is currently no revocation mechanism to help evict these
to make place for the GPU workloads.
(2) A mixed-workload framework (one that runs both GPU and non-GPU workloads)
cannot tell in general if an offer is from an agent with GPUs present, so it
must use attributes to *guarantee* that it does not place non-GPU workloads on
the GPU machine.
The fairness exclusion list allows the operator to ensure that the GPU
allocation does not quickly dominate the share of the role.
> Improve DRF behavior with scarce resources.
> -------------------------------------------
>
> Key: MESOS-5377
> URL: https://issues.apache.org/jira/browse/MESOS-5377
> Project: Mesos
> Issue Type: Epic
> Components: allocation
> Reporter: Benjamin Mahler
> Assignee: Guangya Liu
>
> The allocator currently uses the notion of Weighted [Dominant Resource
> Fairness|https://www.cs.berkeley.edu/~alig/papers/drf.pdf] (WDRF) to
> establish a linear notion of fairness across allocation roles.
> DRF behaves well for resources that are present within each machine in a
> cluster (e.g. CPUs, memory, disk). However, some resources (e.g. GPUs) are
> only present on a subset of machines in the cluster.
> Consider the behavior when there are the following agents in a cluster:
> 1000 agents with (cpus:4,mem:1024,disk:1024)
> 1 agent with (gpus:1,cpus:4,mem:1024,disk:1024)
> If a role wishes to use both GPU and non-GPU resources for tasks, consuming 1
> GPU will lead DRF to consider the role to have a 100% share of the cluster,
> since it consumes 100% of the GPUs in the cluster. This framework will then
> not receive any other offers.
> Among possible improvements, fairness can have understanding of resource
> packages. In a sense there is 1 GPU package that is competed on and 1000
> non-GPU packages competed on, and ideally a role's consumption of the single
> GPU package does not have a large effect on the role's access to the other
> 1000 non-GPU packages.
> In the interim, we should consider having a recommended way to deal with
> scarce resources in the current model.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
