In order to build the executors you will need to build or obtain a mesos
egg and place it in a directory named third_party at the root of the aurora
repo. The filename looks like mesos.native-*-linux-x86_64.egg. It is
produced by default by building mesos from source. You might also want to
have a
You are most likely missing a mesos egg for your platform. See related
thread [1] on how to workaround this problem.
Thanks,
Maxim
[1] -
http://mail-archives.apache.org/mod_mbox/aurora-dev/201410.mbox/%3CCAHD-6f8PkS84Fdp_Y3gnzZtk=TKgmV=02kh4e+5akm6ozdr...@mail.gmail.com%3E
On Tue, Dec 9, 2014 a
I've realized I was using the sample configuration which has these lines
# TODO(Kevin Sweeney): Point these to real URLs.
-thermos_executor_path=/dev/null
-gc_executor_path=/dev/null
So I tried to build the thermos executor with
./pants src/main/python/apache/aurora/executor/bin:thermos_e
This appears suspect:
Fetching URI '/dev/null'
On the scheduler, what's the value supplied for the
--thermos_executor_path command
line argument?
-=Bill
On Mon, Dec 8, 2014 at 12:31 PM, Larry Weya wrote:
> I built Aurora from source and I'm running Mesos 0.21.0
>
> While trying to run the
I built Aurora from source and I'm running Mesos 0.21.0
While trying to run the Tutorial's job I get TASK_LOST from Mesos and
*THROTTLED :
Rescheduled, penalized for 6 ms for flapping* from Aurora
*Here is what I get within the job's stderr*
WARNING: Logging before InitGoogleLogging() is wri
