by

Where communities thrive


  • Join over 1.5M+ people
  • Join over 100K+ communities
  • Free without limits
  • Create your own community
People
Repo info
Activity
Shingo Omura
@everpeace

Thanks for telling so. I'm glad to hear that too.

by the way, I notice just now though. marathon-0.7.3 seems to have some isssues on GUI, doesn't it? I can't see marathon gui when $vagrant up --provider=virtualbox(standalone).

my chrome's dev console reported that /dist/main.js and /dist/main.css returns forbidden 404.
josemolmomillan
@josemolmomillan
for me it's working. but i have an issue with first deployed contained
i get to run the container, mesos return 31000 and 31001 ports; i have added these ports to security group, but when i try to connect to the ports from browser or wget i get connection refused
do you figure something out?
Shingo Omura
@everpeace
did you make sure that your container exposes some ports appropriately or uses host network??
josemolmomillan
@josemolmomillan
Later I pass testing for deploying containers in AWS standalone mode.
Now, i have begun to create the same in a specific VPC 172.30.0.0/16 and several subnets 172.30.[0-3].0/24. ZK node is created perfectly in 172.30.0.0, but when master is creating i got the following error InvalidParameterValue => Address 172.30.1.11 does not fall within the subnet's address range. The first subnet in cluster is 172.30.0.0, but the rest of nodes are configured in otrher subnets like your example. how must i indicate this to vagrant?
Shingo Omura
@everpeace
sorry for inconvenient, for now, vagrant-mesos doesn't support deployment across multi subnets.
so, You have to create a larger subnet, e.g. 172.30.0.0/20 or 172.30.0.0/16.
josemolmomillan
@josemolmomillan
ok. thanks. then which is your advice for zk_ipbase : "172.30.0."
master_ipbase: "172.30.1."
slave_ipbase : "172.30.2."
marathon_ipbase: "172.30.1."
172.3.0.11, 172.30.0.12, 172.30.13, ...?
Shingo Omura
@everpeace
I don't see any problem your setting if you have a subset which is large enough.
Or, if you wanted more flexibility on that, you can modify https://github.com/everpeace/vagrant-mesos/blob/master/multinodes/lib/gen_node_infos.rb
josemolmomillan
@josemolmomillan
thanks. i have updated it and it works better.
nevertheless, which HVM AMI are you using in VPC mode? mesosphere images? some of yours?
josemolmomillan
@josemolmomillan
in VPC mode paravirtual AMIs doesnt work
Shingo Omura
@everpeace
do you mean ami-28ab1640doesn't work in VPC??
mesosphere images? some of yours?
in default setting, vagrant-mesos uses these images: https://github.com/everpeace/packer-mesos#aws
josemolmomillan
@josemolmomillan
in us-east-1 i havent paravirtual machines. only HVM machines
Shingo Omura
@everpeace
I'm not faimiliar with paravirtual/HVM differences...
so, you mean vagrant-mesos doesn't support us-east-1 region because us-east-1 doesn't support only HVM.
josemolmomillan
@josemolmomillan
i think so. the image is paravirtual. the only similar HVM image is from mesosphere
Shingo Omura
@everpeace
actually, these images were build packer with packer-mesos
I'm really glad that you would file PR for that :-)
base image ami is ami-73717d36
josemolmomillan
@josemolmomillan
Ewan Higgs
@ehiggs
For chefdk, do we really need to fill out information to get this?
Ewan Higgs
@ehiggs
bah, "get chef" isn't the place to get chefdk. apparently one clicks an OS logo for a distro similar to theirs.
Shingo Omura
@everpeace
What did you mean? Is your issue resoleved already?
If you found some misleading guidance, please feel free to open an issue :-)
Shingo Omura
@everpeace
Oh, I now see the situation you faced. You are blocked by some input form on the chef site, don't you.
Alexey Kupershtokh
@AlexeyKupershtokh
Hello guys. Does anyone knows if the vagrant-mesos spins up haproxy and the bridge as well? Couldn't find anything related in the running VM nor in the repo files.
Or one is supposed to install it manually?
Alexey Kupershtokh
@AlexeyKupershtokh
BTW, when I vagrant halt the cluster, and then vagrant up again, marathon vm starts, but the marathon process within - does not.
Shingo Omura
@everpeace
sorry for late reply. vagrant up doesn't provision when vm was already initialized. please try vagrant up --provision to execute provision phase again.

if the vagrant-mesos spins up haproxy and the bridge as well?

maybe you can edit Berksfile to insall haproxy and configure it.

Jan
@hanskoff
hi, I'm searching for mesos booting script, and can't find it's address in vagrantfile
Jan
@hanskoff
so I mean I see add_recipe statements
Shingo Omura
@everpeace
you can refer to https://github.com/everpeace/cookbook-mesos to configure mesos-receipie
Shingo Omura
@everpeace

@hanskoff I now see what you meant. I wrote an answer.

https://github.com/everpeace/vagrant-mesos/issues/67#issuecomment-113335682

Keshava
@keshavab
@everpeace On reload --provision, marathon and chronos are not running? Do we need to run them manullay?
Mike Key
@milehimikey
Same question as keshavab
Shingo Omura
@everpeace
bash provision script seems to contain some bug.... I can't spare some time to fix it so far. I'm looking forward to your PR!!
Ammar
@Arwany
Hi, I was able to vagrant up this, it is really helpful. Anyway I need to launch a private registry along, and I have little experience with vagrant, any help?
rajabhathor
@rajabhathor
Ok I'm new to Mesos - I've got the cluster up and running where do I go from here to run Spark apps?
Thanks!
Shingo Omura
@everpeace
I haven't tried to run Spark apps on my vagrant-mesos environment. But I think you would go to public document of Spark: http://spark.apache.org/docs/latest/running-on-mesos.html
guram96
@guram96
hi guys