Chronos keeps failing
@maasg Chronos fails with the following error
[scallop] Error: Excess arguments provided: '/usr/bin/chronos run_jar --http_port 4400 --zk_hosts 172.17.0.3:2181 --master zk://172.17.0.3:2181/mesos --http_port 31309'
Any idea?
The option --http_port appears twice in the command above (4400, 31309). How are you submitting it?
See https://github.com/ContainerSolutions/minimesos/blob/master/opt/apps/chronos.json.
It seems Mesos adds an extra --http_port with the port from the configured port range.
I think the port 4400 port should be removed from the Marathon file. We don't know in advance which port it will get assigned.
Something changed in the chronos port exposure (PORT0 > HTTP API, and `PORT1 > libprocess), recently I reinstated chronos on our cluster and opted for the recently released chronos 3 with this:
echo "---> start chronos"
docker run -d \
-e PORT0=4400 \
-e PORT1=8081 \
--name chronos --restart always \
-p 4400:4400 \
-p 8081:8081 \
mesosphere/chronos:v3.0.1 \
--zk_hosts=${MESOS_MASTER_1}:2181 \
--master=zk://${MESOS_MASTER_1}:2181/mesos \
--hostname=${MESOS_MASTER_1} \
--mesos_role=private \
--mesos_framework_name=chronos
Snippet from https://mesos.github.io/chronos/docs/
To run Chronos with Docker, you must have 2 ports available: one for the HTTP API, and one for libprocess. You must export these ports as environment variables for Chronos to start. For example:
$ docker run --net=host -e PORT0=8080 -e PORT1=8081 mesosphere/chronos:v3.0.0 --zk_hosts 192.168.65.90:2181 --master zk://192.168.65.90:2181/mesos
Your previous link to the chronos blob no longer works, so I can't verify.
Hope this helps @frankscholten and you can add chronos to minimesos
Thanks for the heads up @blijblijblij, I will have a look soon.