+1 (binding)
Tested on CI for centos5/6
@vinodkone
> On Sep 23, 2015, at 3:45 PM, Adam Bordelon wrote:
>
> +1 (binding) Tested on CI for CentOS7, Fedora22, and Ubuntu 14.04.
>
>> On Wed, Sep 23, 2015 at 10:25 AM, haosdent wrote:
>>
>> +1 test on Ubuntu 14.04
>>
>>> On Tue, Sep 22, 2015 at
Created a ticket for us to continue the discussion:
https://issues.apache.org/jira/browse/MESOS-3507
We can try to capture the explicit use-case from Aaron and maybe create
another ticket to track a more-or-less generic path we could go down.
Cheers,
Niklas
On 23 September 2015 at 15:55, Sharma
Discussing in a separate place/JIRA ticket sounds good.
Basically, representing contention using a summary of pending resource
requests from each framework could be the hints to mesos master. However,
this gets into intricacies, not the least of which is diversity of resource
requests, qualified by
+1 (binding) Tested on CI for CentOS7, Fedora22, and Ubuntu 14.04.
On Wed, Sep 23, 2015 at 10:25 AM, haosdent wrote:
> +1 test on Ubuntu 14.04
>
> On Tue, Sep 22, 2015 at 9:06 AM, Adam Bordelon wrote:
>
>> Hi friends,
>>
>> Please vote on releasing the following candidate as Apache Mesos 0.23.1
I'd love to see this solved in a general way; "How does the framework
communicate (insert intent, metric, hint, etc) to mesos".
In one way, the 'webui_url' of in the framework info conveys "This is how
you get to my web ui". As providing a webui was a common pattern for the
frameworks.
This could
Hi Aaron,
You might consider trying to solve the autoscaling problem with Relay, a
Python tool I use to solve this problem. Feel free to shoot me an email if
you are interested.
github.com/sailthru/relay
Alex
On Wed, Sep 23, 2015, 11:03 AM David Greenberg
wrote:
> In addition, this technique
+1 (binding) Tested on CI for CentOS7, Fedora22, and Ubuntu 14.04.
On Tue, Sep 22, 2015 at 1:41 PM, Niklas Nielsen
wrote:
> +1 (binding)
>
> On 21 September 2015 at 11:46, Vinod Kone wrote:
>
>> +1 (binding)
>>
>> Tested on CI for CentOS5 and CentOS6.
>>
>> On Fri, Sep 18, 2015 at 6:21 PM, Adam
Can you run the slave and executor with GLOG_v=1 set on the environment and
try to provide some more context for this error:
> mesos-containerizer: error while loading shared libraries:
> libmesos-0.24.0.so: cannot open shared object file: No such file or
> directory
Are there any logs on the sla
We are adding better support for systemd in 0.25. The ticket is MESOS-3425.
Naturally this is still somewhat experimental, but we would love your
feedback.
We will add some documentation on recommended setups on systemd.
With the changes going into 0.25 you should be able to launch your slave
with
There is a plan for event subscription, but it is still in the early design
phase.
In 0.25 we are adding slave exit hooks: MESOS-3015
This will allow you to generate whatever events you like based on removal
of a slave. This is your best bet in terms of an immediate solution :-)
@Kapil and @Nikla
+joris
On Thu, Sep 17, 2015 at 6:44 AM, tommy xiao wrote:
> read many more report on SSL。 does it mean currently the mesos can't
> support ssl interconn?
>
> 2015-09-17 18:55 GMT+08:00 Carlos Sanchez :
>
>> I got back to SSL and made some progress, SSL is enabled now (I think
>> I needed to expo
I believe some of the contributors from Mesosphere have been thinking about
it, but not sure on the plans. I'll let them reply here.
On Wed, Sep 16, 2015 at 11:11 AM, Paul Bell wrote:
> Thank you, Benjamin.
>
> So, I could periodically request the metrics endpoint, or stream the logs
> (maybe vi
+1 test on Ubuntu 14.04
On Tue, Sep 22, 2015 at 9:06 AM, Adam Bordelon wrote:
> Hi friends,
>
> Please vote on releasing the following candidate as Apache Mesos 0.23.1.
>
> 0.23.1 is a bug fix release and includes the following:
>
> ---
Thank you for the issue link. I will go through to understand which
configuration changes can be done with and without recovery.
On Wed, Sep 23, 2015 at 4:25 PM, Vinod Kone wrote:
> It's not yet possible to make certain slave configuration changes while
> making recovery (reconnecting with old e
Hi Clarke,
Yes, you are right, the mesos is now using 1024 * CPUS as cpushare, you can
refer to
https://github.com/apache/mesos/blob/master/src/docker/docker.cpp#L394 for
detail.
Regarding to cpus, I think that by default the container have same number
of cores/cpus as the docker server, not sure
It's not yet possible to make certain slave configuration changes while
making recovery (reconnecting with old executors) work.
See https://issues.apache.org/jira/browse/MESOS-1739 and attached tickets
for details.
On Wed, Sep 23, 2015 at 7:37 AM, Pradeep Chhetri <
pradeep.chhetr...@gmail.com> wr
In addition, this technique could be implemented in the allocator with an
understanding of global demand: https://www.youtube.com/watch?v=BkBMYUe76oI
That would allow for tunable fair-sharing based on DRF-principles.
On Wed, Sep 23, 2015 at 10:59 AM haosdent wrote:
> Feel free to open a story i
I think you misunderstand. I'm not looking to change the container resource
allocation, I'm looking to have the program in the container make adjustments
(changing static buffer allocations based on memory limits or adjusting the
number of threads based on cpu allocation). But I need a way to co
Feel free to open a story in jira if you think you ideas are awesome. :-)
On Sep 23, 2015 10:54 PM, "Sharma Podila" wrote:
> Ah, OK, thanks. Yes, Fenzo is a Java library.
>
> It might be a nice addition to Mesos master to get a global view of
> contention for resources. In addition to autoscaling
>Is there a way for a process in the docker container to determine how
many cpus it has been allocated?
Because docker limit the resource before start the container, I think there
isn't a way to change container resource usage after it has already run.
You only could allocate resource before launc
Ah, OK, thanks. Yes, Fenzo is a Java library.
It might be a nice addition to Mesos master to get a global view of
contention for resources. In addition to autoscaling, it would be useful in
the allocator.
On Wed, Sep 23, 2015 at 7:29 AM, Aaron Carey wrote:
> Thanks Sharma,
>
> I was in the au
Thank you for the replies.
Paul, I am talking the about the same directory. There is a file named
slave.info inside /tmp/mesos/meta/slaves/latest and this needs to be
cleaned before starting mesos slave with a configuration change.
No i am not using systemd. It is basically sysvinit which is spaw
That's just the scheduler...I did locate the appropriate docker run call and it
appears that mesos cpu offer * 1024 is used as the cpu share so there's no hard
enforcement of total cpus but the scheduler block is scaled to the number of
requested cpus. It also appears there's no way to determine
Thanks Sharma,
I was in the audience for a talk you did about Fenzo at MesosCon :) It looked
great but we're a python shop primarily so the Java requirement would be a
problem for us.
The scaling in the scheduler makes total sense, (obvious when you think about
it!), I was naively hoping for s
Jobs/tasks wait in framework schedulers, not mesos master. Autoscaling
triggers must come from schedulers, not only because that's who knows the
pending task set size, but, also because it knows how many of them need to
be launched right away, on what kind of machines.
We built such an autoscaling
Hi Clarke,
You can take a look at the framework for docker here
https://github.com/apache/mesos/blob/master/src/examples/docker_no_executor_framework.cpp
All of the resource limitations for a docker task should be defined in the
framework.
Thanks,
Guangya
On Wed, Sep 23, 2015 at 9:33 PM, Clark
That's my fallback but it would require modifying and rebuilding the docker
containers and it won't redirect and stderr from the os instance in the
container, just the application that's executing.
From: haosdent [haosd...@gmail.com]
Sent: Wednesday, September 23
When a framework accepts an offer and starts a docker task, how does mesos
enforce the task's allocated cpus? Is CPU share used and scaled appropriately?
Are cpu sets explicitly specified to limit execution to the allocated cpus? Is
there a way for a process in the docker container to determine
How about use 2>&1 ?
On Wed, Sep 23, 2015 at 9:16 PM, Clarke, Trevor wrote:
> Is there a way to have mesos combine stderr with stdout in a docker
> process so there's only one log stream with proper interleave?
>
>
> This message and any enclosures are intended only for the addressee.
> Please
>
Is there a way to have mesos combine stderr with stdout in a docker process so
there's only one log stream with proper interleave?
This message and any enclosures are intended only for the addressee. Please
notify the sender by email if you are not the intended recipient. If you are
not the
I believe Brian might be referring to the "KillMode" in the systemd unit
file:
# the default is cgroup, which means kill all processes
# in the control group of this process, which is not
# what you'd want
KillMode=process
On Wed, Sep 23, 2015 at 8:11 AM, Brian Devins wrote:
> Are you using sys
Are you using systemd? There is a known issue with slave recovery on
systemd. I'm on mobile or I would link you to the last thread around this
but there is a line you can add to the config that is supposed to fix it.
Whether it will fix it is another matter. I am fighting this issue at work
myself.
No, I basically had the same question as Jim (but maybe didn't word it so well
;))
I'll have a look at your response there :)
From: haosdent [haosd...@gmail.com]
Sent: 23 September 2015 10:12
To: user@mesos.apache.org
Subject: Re: Metric for tasks queued/waiting?
Hi Pradeep,
Perhaps I am speaking to a slightly different point, but when I change
/etc/default/mesos-slave to add a new attribute, I have to remove file
/tmp/mesos/meta/slaves/latest.
IIRC, mesos-slave itself, in failing to start after such a change, tells me
to do this:
rm -f /tmp/mesos/meta/s
Hi, @James As far as I know, when the framework request offer,
master/messages_resource_request would increment. And when the framework
accept the offer and launch tasks, master/messages_launch_tasks would
increment. But mesos don't know how many tasks pending in framework,
because I think this inf
Does /metrics/snapshot not satisfy your requirement?
On Wed, Sep 23, 2015 at 4:50 PM, Aaron Carey wrote:
> Hi all,
>
> Is there any way to get a metric of all tasks currently waiting/queued in
> Mesos (across all schedulers)? The snapshot metrics seem to cover ever
> other kind of task state? Th
Hi all. It appears there is a glaring omission in the 'Tasks' section of
the following doc;
http://mesos.apache.org/documentation/latest/monitoring/
Shouldn't there be a 'Tasks waiting' metric!? We generally have tasks
hanging around for
a while because their resource requests can't (yet) be met
Hi all,
Is there any way to get a metric of all tasks currently waiting/queued in Mesos
(across all schedulers)? The snapshot metrics seem to cover ever other kind of
task state? This would be quite useful for auto-scaling purposes..
Thanks,
Aaron
Hello all,
I have often faced this problem that whenever i try to add some
configuration parameter to mesos-slave or change any configuration (eg. add
a new attribute in mesos-slave), the mesos slave doesnt come up on restart.
I have to delete the slave.info file and then restart the slave but it
39 matches
Mail list logo
