Open punit-kulal opened 3 years ago
Yes, this feature is very helpful. When using one of the proxy nodes to get cluster information, we want to be able to return a list of available nodes for the entire Kafka-REST cluster, which is necessary for fault tolerance of clients using the proxy to write messages. I want to add nodes to the following data nodes like this:
{
"kind": "KafkaClusterList",
"metadata": {
"self": "http://localhost:9391/v3/clusters",
"next": null
},
"data": [
{
"kind": "KafkaCluster",
"metadata": {
"self": "http://localhost:9391/v3/clusters/cluster-1",
"resource_name": "crn:///kafka=cluster-1"
},
"cluster_id": "cluster-1",
"nodes": [
"http://192.168.1.1:9391/",
"http://192.168.1.2:9391"
],
"controller": {
"related": "http://localhost:9391/v3/clusters/cluster-1/brokers/1"
},
"acls": {
"related": "http://localhost:9391/v3/clusters/cluster-1/acls"
},
"brokers": {
"related": "http://localhost:9391/v3/clusters/cluster-1/brokers"
},
"broker_configs": {
"related": "http://localhost:9391/v3/clusters/cluster-1/broker-configs"
},
"consumer_groups": {
"related": "http://localhost:9391/v3/clusters/cluster-1/consumer-groups"
},
"topics": {
"related": "http://localhost:9391/v3/clusters/cluster-1/topics"
},
"partition_reassignments": {
"related": "http://localhost:9391/v3/clusters/cluster-1/topics/-/partitions/-/reassignment"
}
}
]
}
Yes, this feature is very helpful. When using one of the proxy nodes to get cluster information, we want to be able to return a list of available nodes for the entire Kafka-REST cluster, which is necessary for fault tolerance of clients using the proxy to write messages. I want to add nodes to the following data nodes like this:
{ "kind": "KafkaClusterList", "metadata": { "self": "http://localhost:9391/v3/clusters", "next": null }, "data": [ { "kind": "KafkaCluster", "metadata": { "self": "http://localhost:9391/v3/clusters/cluster-1", "resource_name": "crn:///kafka=cluster-1" }, "cluster_id": "cluster-1", "nodes": [ "http://192.168.1.1:9391/", "http://192.168.1.2:9391" ], "controller": { "related": "http://localhost:9391/v3/clusters/cluster-1/brokers/1" }, "acls": { "related": "http://localhost:9391/v3/clusters/cluster-1/acls" }, "brokers": { "related": "http://localhost:9391/v3/clusters/cluster-1/brokers" }, "broker_configs": { "related": "http://localhost:9391/v3/clusters/cluster-1/broker-configs" }, "consumer_groups": { "related": "http://localhost:9391/v3/clusters/cluster-1/consumer-groups" }, "topics": { "related": "http://localhost:9391/v3/clusters/cluster-1/topics" }, "partition_reassignments": { "related": "http://localhost:9391/v3/clusters/cluster-1/topics/-/partitions/-/reassignment" } } ] }
Sorry, the previous data format is wrong, this is a new data format, put nodes information in ClusterDataList, I don't know what you think.
{
"kind": "KafkaClusterList",
"metadata": {
"self": "http://localhost:9391/v3/clusters",
"next": null
},
"nodes": {
"self": "http://localhost:8082/",
"available_node": [
{
"host": "http://192.168.1.1:8082/",
"healthy": true,
"weight": 1.0,
"enabled": true,
"metadata": {
"foo1": "bar1"
}
},
{
"host": "http://192.168.1.2:8082/",
"healthy": true,
"weight": 1.0,
"enabled": true,
"metadata": {
"foo2": "bar2"
}
}
]
},
"data": [
{
"kind": "KafkaCluster",
"metadata": {
"self": "http://localhost:9391/v3/clusters/cluster-1",
"resource_name": "crn:///kafka=cluster-1"
},
"cluster_id": "cluster-1",
"controller": {
"related": "http://localhost:9391/v3/clusters/cluster-1/brokers/1"
},
"acls": {
"related": "http://localhost:9391/v3/clusters/cluster-1/acls"
},
"brokers": {
"related": "http://localhost:9391/v3/clusters/cluster-1/brokers"
},
"broker_configs": {
"related": "http://localhost:9391/v3/clusters/cluster-1/broker-configs"
},
"consumer_groups": {
"related": "http://localhost:9391/v3/clusters/cluster-1/consumer-groups"
},
"topics": {
"related": "http://localhost:9391/v3/clusters/cluster-1/topics"
},
"partition_reassignments": {
"related": "http://localhost:9391/v3/clusters/cluster-1/topics/-/partitions/-/reassignment"
}
}
]
}
Could you point me to any open apache wiki issue for adding multiple cluster support. Would be happy to contribute