As per the documentation, it's not supported yet. OpenStack Supported Versions Cloudbreak was tested against the following versions of Red Hat Distribution of OpenStack (RDO): Juno Kilo Liberty Mitaka ... View more

You have more information in the logs and Cloudformation template - nevertheless see this: http://docs.aws.amazon.com/AmazonVPC/latest/UserGuide/VPC_Appendix_Limits.html - the The maximum number of addresses has been reached means that on AWS you are hitting you account limits. You should ask AWS to increase it. ... View more

Yes, we did it - we will come up with an example next week, stay tuned. Feel free to go ahead with your thoughts on this, however this is what is coming: 1. Authentication from LDAP/AD 2. LDAP/AD group mappings to UAA scopes - as you know in Cloudbreak every resource has a scope - thus based on your LDAP/AD settings you can allow/restrict users to operations on resources 3. Visibility of resources based on LDAP/AD groups - this requires some code changes on our side and we are working on this currently. For 1-2 you can already do that by creating an UAA mapping table and generate the groups to UAA scopes Copy the group names to a file (groups.txt): select displayname from groups; (remove first space chars and empty lines) Then you can use sigil to generate the SQL inserts for the external group mapping table. Save this into a file (sigil.template): {{ range $k, $v := stdin|split "\n"}} INSERT INTO external_group_mapping (group_id, external_group, added, origin) VALUES ((select id from groups where displayname='{{$v}}'), 'cn=admin,ou=scopes,dc=ad,dc=seq,dc=com', '2016-09-30 19:28:24.255', 'ldap');{{end}} then: cat groups.txt | sigil -f sigil.template You can get the sigil tool from https://github.com/gliderlabs/sigil. ... View more

Are you using a custom blueprint or yours? ... View more

There is no one config - fits all setup. Like on any public providers where you have pre-set configs, in OpenStack you can create your own instance configs. Since Cloudbreak supports heterogenous clusters I suggest to create different Nova instances which fits best your different HDP components or workload (e.g. high memory for Spark, multi-core for compute heavy workloads, etc). Re OpenStack version - for now we support stock Kilo and Juno. I suggest to check the docs: http://sequenceiq.com/cloudbreak-docs/latest/openstack/ If you don't have an OpenStack cluster yet you might consider checking these Ansible scripts: https://github.com/sequenceiq/openstack-ansible ... View more

In general if you check the default blueprints in Cloudbreak you will see that none of the Master hostgroups are doing any work in case you submit a job/query.etc - they only run master/server and UI components. All the work is done by the slave/worker hostgroups - where you have the datanodes or for example HBase region servers. ... View more

I assume you'd like to upscale a worker hostgroup - not a hostgroup which contains master components (like HIVE_SERVER) in 'host_group_master_2'. Master hostgroups can't be upscaled - if you'd like to do HA (for components which support that) than it's a different story. Bottom line - you can upscale worker hostgroups, not master ones (as it does not make too much sense upscale master hostgroups anyway). ... View more

They should not be opened globally - they should be accessed by the Cloudbreak application. You can deploy the Cloudbreak application (with CBD) on the same VPC (and/or subnet) as where the HDP clusters are provisioned - in that case it should not be opened globally. If you use the hosted Cloudbreak application (cloudbreak.sequenceiq.com) that you will have to open globally - but it's not really recommended, as you should use your own Cloudbreak instance. ... View more

All resources created by Cloudbreak can be modified on AWS site - for AWS makes no difference from where the API call or Cloudformation template is made, nor differentiates it whatsoever - so you either using a different account or IAM role as security groups created by Cloudbreak can be modified after. People often do this and open or close ports on long running clusters at the time it's needed. It's in the documentation - IMPORTANT 443 and 22 ports needs to be there in every security group otherwise Cloudbreak won't be able to communicate with the provisioned cluster - http://sequenceiq.com/cloudbreak-docs/latest/aws/#infrastructure-templates under Security groups. ... View more

For the networks we don't really care - we support all these scenarios: Create a new network and a new subnet Create a new subnet in an existing network Use an existing subnet in an existing network How you configure your existing network it's up to you. For volumes we do do it through API or Heat templates and there is no out of the box customization available yet. You might want to do the LUKS through Cinder volume type creates and modify the template/code in Cloudbreak to reuse that encrypted volume. For custom things (e.g. web server) - currently it's the default Ambari stack with all the OTB components/services - though Ambari supports custom stacks. ... View more

Can you retry now on the hosted version - it should work now! Also if you are using CBD than you should do the following: In your Profile export DOCKER_TAG_CLOUDBREAK=1.2.6-rc.3 And then restart CBD with: cbd kill && cbd regenerate && cbd start ... View more