See also https://issues.apache.org/jira/browse/MESOS-155 for preemption
and the inverse offer mechanism in
https://issues.apache.org/jira/browse/MESOS-1474 (Maintenance Primitives)
On Tue, Feb 24, 2015 at 10:00 AM, Vinod Kone vinodk...@apache.org wrote:
Not currently, no. But we've definitely
Hi,
I'm new to Mesos technology. Browsing thru the documentation, I couldn't
find how a running task can be killed by Mesos master (/ its pluggable
resource allocation module), or by a framework on top of Mesos. Is it
possible with current Mesos tools, or a task stops only when it is
finished? If
There's a killTask() call frameworks can use.
@vinodkone
On Feb 24, 2015, at 7:42 AM, dev middleware dev.middlew...@gmail.com wrote:
Hi,
I'm new to Mesos technology. Browsing thru the documentation, I couldn't find
how a running task can be killed by Mesos master (/ its pluggable
Adam,
Thanks for the pointer, I was able to pull the logs for the docker run
command. Up to my understanding it is actually pulling the image, but it is
having trouble starting the actual docker, I highlighted in red what I
think is the main reason for the error: a bad format for the volume where
Hi Anirudha,
we are currently using mesos-dns and latest version compiles just fine
(ubuntu 14.04, go 1.2.1).
2015-02-24 18:36 GMT+01:00 Ken Sipe kens...@gmail.com:
Anirudha,
Did you follow: http://mesosphere.github.io/mesos-dns/docs/ ?
the build should work according to the build
Thanks. Can this also be done by the master (resource allocation module)?
It might be required to kill a task belonging to a misbehaving framework.
Or to implement a centrally enforced policy that can stop low priority jobs
from one framework, in order to make room for high priority jobs from
Whats the plan for mesos DNS? The dns lib is not even released.
even the build fails with syntax errors.
is there a particular way to get this working?
--
Anirudha
--
sudo go build -o mesos-dns
# github.com/miekg/dns
/usr/lib/go/src/pkg/github.com/miekg/dns/msg.go:1936: syntax error:
Ah, colons in the executorId. What version of Mesos are you running? You
might be hitting https://issues.apache.org/jira/browse/MESOS-1833
On Tue, Feb 24, 2015 at 9:39 AM, max square max2subscr...@gmail.com wrote:
Adam,
Thanks for the pointer, I was able to pull the logs for the docker run
Anirudha,
Did you follow: http://mesosphere.github.io/mesos-dns/docs/
http://mesosphere.github.io/mesos-dns/docs/ ?
the build should work according to the build instructions.
ken
On Feb 24, 2015, at 11:31 AM, Anirudha Jadhav aniru...@nyu.edu wrote:
Whats the plan for mesos DNS? The dns lib
Not currently, no. But we've definitely discussed these semantics (
https://issues.apache.org/jira/browse/MESOS-354). The timeline for this
feature is not clear yet.
On Tue, Feb 24, 2015 at 9:23 AM, dev middleware dev.middlew...@gmail.com
wrote:
Thanks. Can this also be done by the master
sounds good, tried debugging again, I had the golang installed from apt-get
which got an older version of GO.
All is good now.
is there a timeline for this to be released.
Happy to see this is an active project!
-Ani
On Tue, Feb 24, 2015 at 12:43 PM, Ondrej Smola ondrej.sm...@gmail.com
wrote:
It appears your configuration is off… as you suspected.. the master
registration should NOT be 127.0.0.1 or 127.0.1.1.For each master if you
configure the IP in a file named ip under `/etc/mesos-master` you should be
good (after restarting the master)
my configurations under
Hi Devin,
I am new to Mesos as well, and I just configured it had the same problem
like yours.
For your reference, what my fix was use the actually master IP instead,
then slave will pick it up and connected. I really wonder if 127.0.0.1,
then Slave will use it to connect itself and that is why
Adam/Tim,
that's exactly the issue thanks! I am using 0.20.0 currently.
Is there any particular suggested update approach?
FYI I have another in-house cluster running 0.21.1 and it worked fine.
Haven't been able to get that to work on AWS though.
Thanks!
Sergio Daniel
On Tue, Feb 24, 2015 at
I would like to point out that using a new FrameworkID is not a solution to
this problem. This means that a cluster operator has to drain the entire
cluster to enable checkpointing, or lose all previous tasks. Both scenarios
are not desirable.
Fortunately it is possible to do this without
Changing FrameworkInfo (while keeping the FrameworkID) is not handled
correctly by Mesos at the moment. This is what you currently need to do to
propagate FrameworkInfo.checkpoint throughout the cluster.
-- Update FrameworkInfo inside your framework and re-register with master.
(Old FrameworkInfo
I setup a vanilla basic cluster - but also put docker on the mesos slaves -
and ran marathon (I used the mesophere packages for ubuntu).
I noted using the default settings of the mesos containerizer I could
access the docker daemon, run containers etc - which surprised me - is this
expected? I
Hey folks,
Is there a best practice for rolling out FrameworkInfo changes? We need to
set checkpoint to true, so I redeployed our framework with the new settings
(with tasks still running), but when I hit a slave's stats.json endpoint,
it appears that the old FrameworkInfo data is still there
Mesos checkpoints the FrameworkInfo into disk, and recovers it on relaunch.
I don't think we expose any API to remove the framework manually though if
you really want to keep the FrameworkID. If you hit the failover timeout
the framework will get removed from the master and slave.
I think for
19 matches
Mail list logo
