You are viewing a plain text version of this content. The canonical link for it is here.
Posted to dev@ambari.apache.org by "John Speidel (JIRA)" <ji...@apache.org> on 2014/03/13 23:05:55 UTC
[jira] [Updated] (AMBARI-5077) Provision a cluster from a an Ambari
Blueprint
[ https://issues.apache.org/jira/browse/AMBARI-5077?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]
John Speidel updated AMBARI-5077:
---------------------------------
Description:
Allow a cluster to be fully provisioned via the Ambari REST API with a single REST API call by specifying a blueprint name.
In addition to the blueprint name, host information for each host group being deployed should be specified.
The process to provision a cluster via the REST API using blueprints is as follows:
- Create a blueprint resource instance on the Ambari instance which is going to be used to provision the cluster. The blueprint may have been hand created or exported from a running cluster.
POST myAmbariServer:8080/api/v1/blueprints/single-node-test
{code}
{
"host_groups" : [
{
"name" : "host_group_1",
"components" : [
{
"name" : "HISTORYSERVER"
},
{
"name" : "JOBTRACKER"
},
{
"name" : "NAMENODE"
},
{
"name" : "OOZIE_SERVER"
},
{
"name" : "NAGIOS_SERVER"
},
{
"name" : "SECONDARY_NAMENODE"
},
{
"name" : "AMBARI_SERVER"
},
{
"name" : "GANGLIA_SERVER"
},
{
"name" : "GANGLIA_MONITOR"
},
{
"name" : "DATANODE"
},
{
"name" : "TASKTRACKER"
},
{
"name" : "HDFS_CLIENT"
},
{
"name" : "MAPREDUCE_CLIENT"
},
{
"name" : "OOZIE_CLIENT"
},
{
"name" : "GANGLIA_MONITOR"
}
],
"cardinality" : "1"
}
],
"Blueprints" : {
"blueprint_name" : "single-node-test",
"stack_name" : "HDP",
"stack_version" : "1.3.3"
}
}
{code}
For more information on exporting a blueprint see: https://issues.apache.org/jira/browse/AMBARI-4786 For more information on saving a blueprint see: https://issues.apache.org/jira/browse/AMBARI-4467
- Invoke a POST on the cluster endpoint and specify the blueprint related information. Host information must be provided for each host group being deployed in the cluster. Each host specified will run the components associated with the host group. The host group names specified in the request must match a host group in the specified blueprint.
POST myAmbariServer:8080/api/v1/clusters/c1
{code}
{
"blueprint" : "single-node-test",
"host-groups" :[
{
"name" : "host_group_1",
"hosts" : [
{
"fqdn" : "myHost.novalocal",
"ip" : "172.18.192.3"
}
]
}
]
{code}
Some limitations in this change which will be addressed shortly in other patches:
- configuration information included in the blueprint will not be processed meaning that the cluster will be provisioned with default configurations
- Ambari Agent must be running and have registered with the server on all hosts in the cluster prior to invoking the REST API to provision the cluster
was:
Allow a cluster to be fully provisioned via the Ambari REST API with a single REST API call by specifying a blueprint name.
In addition to the blueprint name, host information for each host group being deployed should be specified.
The process to provision a cluster via the REST API using blueprints is as follows:
- Save a blueprint to the Ambari instance which is going to be used to provision the cluster. The blueprint may be have been hand created or exported from a running cluster.
POST myAmbariServer:8080/api/v1/blueprints/single-node-test
{code}
{
"host_groups" : [
{
"name" : "host_group_1",
"components" : [
{
"name" : "HISTORYSERVER"
},
{
"name" : "JOBTRACKER"
},
{
"name" : "NAMENODE"
},
{
"name" : "OOZIE_SERVER"
},
{
"name" : "NAGIOS_SERVER"
},
{
"name" : "SECONDARY_NAMENODE"
},
{
"name" : "AMBARI_SERVER"
},
{
"name" : "GANGLIA_SERVER"
},
{
"name" : "GANGLIA_MONITOR"
},
{
"name" : "DATANODE"
},
{
"name" : "TASKTRACKER"
},
{
"name" : "HDFS_CLIENT"
},
{
"name" : "MAPREDUCE_CLIENT"
},
{
"name" : "OOZIE_CLIENT"
},
{
"name" : "GANGLIA_MONITOR"
}
],
"cardinality" : "1"
}
],
"Blueprints" : {
"blueprint_name" : "single-node-test",
"stack_name" : "HDP",
"stack_version" : "1.3.3"
}
}
{code}
For more information on exporting a blueprint see: https://issues.apache.org/jira/browse/AMBARI-4786 For more information on saving a blueprint see: https://issues.apache.org/jira/browse/AMBARI-4467
- Invoke a POST on the cluster endpoint and specify the blueprint related information. Host information must be provided for each host group being deployed in the cluster. Each host specified will run the components associated with the host group. The host group names specified in the request must match a host group in the specified blueprint.
POST myAmbariServer:8080/api/v1/clusters/c1
{code}
{
"blueprint" : "single-node-test",
"host-groups" :[
{
"name" : "host_group_1",
"hosts" : [
{
"fqdn" : "myHost.novalocal",
"ip" : "172.18.192.3"
}
]
}
]
{code}
Some limitations in this change which will be addressed shortly in other patches:
- configuration information included in the blueprint will not be processed meaning that the cluster will be provisioned with default configurations
- Ambari Agent must be running and have registered with the server on all hosts in the cluster prior to invoking the REST API to provision the cluster
> Provision a cluster from a an Ambari Blueprint
> ----------------------------------------------
>
> Key: AMBARI-5077
> URL: https://issues.apache.org/jira/browse/AMBARI-5077
> Project: Ambari
> Issue Type: Improvement
> Reporter: John Speidel
> Assignee: John Speidel
> Fix For: 1.5.0
>
>
> Allow a cluster to be fully provisioned via the Ambari REST API with a single REST API call by specifying a blueprint name.
> In addition to the blueprint name, host information for each host group being deployed should be specified.
> The process to provision a cluster via the REST API using blueprints is as follows:
> - Create a blueprint resource instance on the Ambari instance which is going to be used to provision the cluster. The blueprint may have been hand created or exported from a running cluster.
> POST myAmbariServer:8080/api/v1/blueprints/single-node-test
> {code}
> {
> "host_groups" : [
> {
> "name" : "host_group_1",
> "components" : [
> {
> "name" : "HISTORYSERVER"
> },
> {
> "name" : "JOBTRACKER"
> },
> {
> "name" : "NAMENODE"
> },
> {
> "name" : "OOZIE_SERVER"
> },
> {
> "name" : "NAGIOS_SERVER"
> },
> {
> "name" : "SECONDARY_NAMENODE"
> },
> {
> "name" : "AMBARI_SERVER"
> },
> {
> "name" : "GANGLIA_SERVER"
> },
> {
> "name" : "GANGLIA_MONITOR"
> },
> {
> "name" : "DATANODE"
> },
> {
> "name" : "TASKTRACKER"
> },
> {
> "name" : "HDFS_CLIENT"
> },
> {
> "name" : "MAPREDUCE_CLIENT"
> },
> {
> "name" : "OOZIE_CLIENT"
> },
> {
> "name" : "GANGLIA_MONITOR"
> }
> ],
> "cardinality" : "1"
> }
> ],
> "Blueprints" : {
> "blueprint_name" : "single-node-test",
> "stack_name" : "HDP",
> "stack_version" : "1.3.3"
> }
> }
> {code}
> For more information on exporting a blueprint see: https://issues.apache.org/jira/browse/AMBARI-4786 For more information on saving a blueprint see: https://issues.apache.org/jira/browse/AMBARI-4467
> - Invoke a POST on the cluster endpoint and specify the blueprint related information. Host information must be provided for each host group being deployed in the cluster. Each host specified will run the components associated with the host group. The host group names specified in the request must match a host group in the specified blueprint.
> POST myAmbariServer:8080/api/v1/clusters/c1
> {code}
> {
> "blueprint" : "single-node-test",
> "host-groups" :[
> {
> "name" : "host_group_1",
> "hosts" : [
> {
> "fqdn" : "myHost.novalocal",
> "ip" : "172.18.192.3"
> }
> ]
> }
> ]
> {code}
> Some limitations in this change which will be addressed shortly in other patches:
> - configuration information included in the blueprint will not be processed meaning that the cluster will be provisioned with default configurations
> - Ambari Agent must be running and have registered with the server on all hosts in the cluster prior to invoking the REST API to provision the cluster
--
This message was sent by Atlassian JIRA
(v6.2#6252)