Make local hostname (edge2ai-x.dim.local) unique for multi-cluster de…

…ployments.

Before this change all clusters in a multi-cluster deployment would have the same
local hostname (edge2ai-1.dim.local). Now those names are unique per host and have
the format edge2ai-X.dim.local, where X is the cluster ID.

setup/terraform/resources/bootstrap.conf

@@ -36,8 +36,8 @@ nifi.minifi.config=./conf/config.yml
 
 # MiNiFi Command & Control Configuration
 nifi.c2.enable=true
-nifi.c2.rest.url=http://edge2ai-1.dim.local:10088/efm/api/c2-protocol/heartbeat
-nifi.c2.rest.url.ack=http://edge2ai-1.dim.local:10088/efm/api/c2-protocol/acknowledge
+nifi.c2.rest.url=http://LOCAL_HOSTNAME:10088/efm/api/c2-protocol/heartbeat
+nifi.c2.rest.url.ack=http://LOCAL_HOSTNAME:10088/efm/api/c2-protocol/acknowledge
 nifi.c2.agent.heartbeat.period=10000
 nifi.c2.agent.class=iot-1
 nifi.c2.agent.identifier=agent-iot-1

setup/terraform/resources/bootstrap.conf.tls

@@ -36,8 +36,8 @@ nifi.minifi.config=./conf/config.yml
 
 # MiNiFi Command & Control Configuration
 nifi.c2.enable=true
-nifi.c2.rest.url=http://edge2ai-1.dim.local:10088/efm/api/c2-protocol/heartbeat
-nifi.c2.rest.url.ack=http://edge2ai-1.dim.local:10088/efm/api/c2-protocol/acknowledge
+nifi.c2.rest.url=http://LOCAL_HOSTNAME:10088/efm/api/c2-protocol/heartbeat
+nifi.c2.rest.url.ack=http://LOCAL_HOSTNAME:10088/efm/api/c2-protocol/acknowledge
 nifi.c2.agent.heartbeat.period=10000
 nifi.c2.agent.class=iot-1
 nifi.c2.agent.identifier=agent-iot-1

setup/terraform/resources/common.sh

@@ -33,6 +33,20 @@ export TRUSTSTORE_PEM=${SEC_BASE}/x509/truststore.pem
 export KEYSTORE_JKS=${SEC_BASE}/jks/keystore.jks
 export TRUSTSTORE_JKS=${SEC_BASE}/jks/truststore.jks
 
+# Load cluster metadata
+PUBLIC_DNS=${PUBLIC_DNS:-dummy}
+if [[ -f $BASE_DIR/clusters_metadata.sh ]]; then
+  source $BASE_DIR/clusters_metadata.sh
+  PEER_CLUSTER_ID=$(( (CLUSTER_ID/2)*2 + (CLUSTER_ID+1)%2 ))
+  PEER_PUBLIC_DNS=$(echo "$CLUSTERS_PUBLIC_DNS" | awk -F, -v pos=$(( PEER_CLUSTER_ID + 1 )) '{print $pos}')
+  PEER_PUBLIC_DNS=${PEER_PUBLIC_DNS:-$PUBLIC_DNS}
+else
+  CLUSTER_ID=0
+  PEER_CLUSTER_ID=0
+  PEER_PUBLIC_DNS=$PUBLIC_DNS
+fi
+LOCAL_HOSTNAME=edge2ai-${CLUSTER_ID}.dim.local
+export CLUSTER_ID PEER_CLUSTER_ID PEER_PUBLIC_DNS LOCAL_HOSTNAME
 
 function is_kerberos_enabled() {
   echo $ENABLE_KERBEROS
@@ -623,7 +637,7 @@ function create_certs() {
 
   # Create CSR
   local public_ip=$(curl -s http://ifconfig.me || curl -s http://api.ipify.org/)
-  export ALT_NAMES="DNS:edge2ai-1.dim.local,DNS:$(hostname -f),DNS:*.${public_ip}.nip.io,DNS:*.cdsw.${public_ip}.nip.io"
+  export ALT_NAMES="DNS:${LOCAL_HOSTNAME},DNS:$(hostname -f),DNS:*.${public_ip}.nip.io,DNS:*.cdsw.${public_ip}.nip.io"
   openssl req\
     -new\
     -key ${KEY_PEM} \
@@ -662,7 +676,7 @@ EOF
   # Sign cert
   if [[ $ipa_host != "" ]]; then
     kinit -kt $KEYTABS_DIR/admin.keytab admin
-    ipa host-add-principal $(hostname -f) "host/edge2ai-1.dim.local"
+    ipa host-add-principal $(hostname -f) "host/${LOCAL_HOSTNAME}"
     ipa host-add-principal $(hostname -f) "host/*.${public_ip}.nip.io"
     ipa host-add-principal $(hostname -f) "host/*.cdsw.${public_ip}.nip.io"
     ipa cert-request ${CSR_PEM} --principal=host/$(hostname -f)
@@ -778,9 +792,9 @@ EOF
   cat $CERT_PEM $ROOT_PEM >> $sib_cert
   chown shellinabox:shellinabox $sib_cert
   chmod 400 $sib_cert
-  rm -f ${sib_dir}/certificate-{localhost,edge2ai-1.dim.local,${CLUSTER_HOST}}.pem
+  rm -f ${sib_dir}/certificate-{localhost,${LOCAL_HOSTNAME},${CLUSTER_HOST}}.pem
   ln -s $sib_cert ${sib_dir}/certificate-localhost.pem
-  ln -s $sib_cert ${sib_dir}/certificate-edge2ai-1.dim.local.pem
+  ln -s $sib_cert ${sib_dir}/certificate-${LOCAL_HOSTNAME}.pem
   ln -s $sib_cert ${sib_dir}/certificate-${CLUSTER_HOST}.pem
 
 }

setup/terraform/resources/create_cluster.py

@@ -8,11 +8,10 @@
 from datetime import datetime
 from optparse import OptionParser
 import cm_client
-import json
 import os
 import re
 import requests
-import sys
+import socket
 import time
 import urllib3
 
@@ -90,7 +89,7 @@ def cm_version():
 
 def the_pwd():
     return os.environ['THE_PWD']
-    
+
 def cluster_id():
     try:
         if 'CLUSTER_ID' in os.environ:
@@ -100,6 +99,9 @@ def cluster_id():
 
     return 0
 
+def local_hostname():
+    return os.environ.get('LOCAL_HOSTNAME', 'edge2ai-1.local.dim')
+
 class ClusterCreator:
     def __init__(self, host, krb_princ='scm/[email protected]', tls_ca_cert=None):
         self.host = host
@@ -273,7 +275,7 @@ def setup_cm(self, key_file, cm_repo_url, use_kerberos, use_tls, kerberos_type,
 
         # Update cluster banner
         c_id = cluster_id()
-        banner = 'CLUSTER_{}'.format(c_id)
+        banner = 'Cluster ID: {}, Host: {}'.format(c_id, socket.gethostname())
         header_color = HEADER_COLORS[c_id % len(HEADER_COLORS)]
         self.cm_api.update_config(
             message='Customizing CM header and banner',
@@ -339,8 +341,8 @@ def _enable_kerberos(self, kerberos_type, ipa_host):
         ]
         if kerberos_type == 'MIT':
             config += [
-                cm_client.ApiConfig(name='KDC_ADMIN_HOST', value='edge2ai-1.dim.local'),
-                cm_client.ApiConfig(name='KDC_HOST', value='edge2ai-1.dim.local'),
+                cm_client.ApiConfig(name='KDC_ADMIN_HOST', value=local_hostname()),
+                cm_client.ApiConfig(name='KDC_HOST', value=local_hostname()),
                 cm_client.ApiConfig(name='KDC_TYPE', value='MIT KDC'),
             ]
         else:

setup/terraform/resources/labs/utils/postgres.py

@@ -5,8 +5,8 @@
 
 
 def execute_sql(cmd, db_name, username, password):
-    cmd_line = 'PGPASSWORD={pwd} psql --host edge2ai-1.dim.local --port 5432 --username {usr} {db}'.format(
-        usr=username, pwd=password, db=db_name)
+    cmd_line = 'PGPASSWORD={pwd} psql --host {hostname} --port 5432 --username {usr} {db}'.format(
+        usr=username, pwd=password, db=db_name, hostname=get_hostname())
     proc = Popen(cmd_line, shell=True, stdin=PIPE, stdout=PIPE, stderr=PIPE)
     stdout, stderr = proc.communicate(cmd.encode('utf-8'))
     return proc.returncode, stdout, stderr

setup/terraform/resources/labs/utils/ssb.py

@@ -13,6 +13,7 @@
 _API_EXTERNAL = 'external'
 _API_UI = 'ui'
 
+
 def _get_api_url():
     return get_url_scheme() + '://cdp.{}.nip.io:8000/api/v1'.format(get_public_ip())
 
@@ -41,6 +42,8 @@ def _api_call(func, path, data=None, files=None, headers=None, api_type=_API_INT
     if api_type != _API_UI:
         headers['Content-Type'] = 'application/json'
         data = json.dumps(data)
+    if api_type == _API_EXTERNAL:
+        headers['Username'] = 'admin'
     if token:
         headers['X-CSRF-TOKEN'] = _SSB_CSRF_TOKEN
     url = _get_url(api_type) + path
@@ -171,7 +174,6 @@ def execute_sql(stmt, job_name=None, parallelism=None, sample_interval_millis=No
     }
     headers = {
         'Accept': 'application/json',
-        'Username': 'admin',
         'Content-Type': 'application/json',
     }
     return _api_post('/ssb/sql/execute', data, headers=headers, api_type=_API_EXTERNAL)

setup/terraform/resources/labs/workshop_cdc.py

@@ -38,7 +38,7 @@
   name STRING
 ) WITH (
   'connector' = 'postgres-cdc',
-  'hostname' = 'edge2ai-1.dim.local',
+  'hostname' = '{hostname}',
   'username' = 'cdc_user',
   'password' = 'supersecret1',
   'database-name' = 'cdc_test',
@@ -49,7 +49,7 @@
   'debezium.slot.name' = 'flink',
   'debezium.snapshot.mode' = 'initial'
 );
-'''
+'''.format(hostname=get_hostname())
 
 LAB3_TRANSACTIONS = '''
 INSERT INTO transactions
@@ -75,13 +75,13 @@
   PRIMARY KEY (id) NOT ENFORCED
 ) WITH (
   'connector' = 'jdbc',
-  'url' = 'jdbc:postgresql://edge2ai-1.dim.local:5432/cdc_test',
+  'url' = 'jdbc:postgresql://{hostname}:5432/cdc_test',
   'table-name' = 'trans_replica',
   'password' = '{pwd}',
   'username' = 'cdc_user',
   'driver' = 'org.postgresql.Driver'
 );
-'''.format(pwd=get_the_pwd())
+'''.format(pwd=get_the_pwd(), hostname=get_hostname())
 
 LAB4_INSERT_INTO_REPLICA = '''
 INSERT INTO trans_replica
@@ -95,13 +95,13 @@
   `name` VARCHAR(2147483647)
 ) WITH (
   'connector' = 'kafka',
-  'properties.bootstrap.servers' = 'edge2ai-1.dim.local:9092',
+  'properties.bootstrap.servers' = '{hostname}:9092',
   'topic' = 'trans_changelog',
   'key.format' = 'json',
   'key.fields' = 'id',
   'value.format' = 'debezium-json'
 );
-'''
+'''.format(hostname=get_hostname())
 
 LAB5_INSERT_CHANGELOG = '''
 INSERT INTO trans_changelog

setup/terraform/resources/labs/workshop_dataviz.py

@@ -9,9 +9,9 @@
 from .utils import dataviz
 
 CONNECTION_TYPE = 'impyla'
-CONNECTION_NAME = 'Local Impala1'
+CONNECTION_NAME = 'Local Impala'
 CONNECTION_PARAMS = {
-    "HOST": "cdp.52.26.198.174.nip.io",
+    "HOST": get_hostname(),
     "PORT": "21050",
     "MODE": "binary",
     "AUTH": "nosasl",

setup/terraform/resources/labs/workshop_edge.py

@@ -95,7 +95,7 @@ def lab2_edge_flow(self):
             'org.apache.nifi.processors.mqtt.ConsumeMQTT',
             (100, 100),
             {
-                'Broker URI': 'tcp://edge2ai-1.dim.local:1883',
+                'Broker URI': 'tcp://{hostname}:1883'.format(hostname=get_hostname()),
                 'Client ID': 'minifi-iot',
                 'Topic Filter': 'iot/#',
                 'Max Queue Size': '60',

setup/terraform/resources/labs/workshop_ssb.py

@@ -7,7 +7,7 @@
 from .utils import ssb, schreg
 
 KAFKA_PROVIDER_NAME = 'edge2ai-kafka'
-KAFKA_PROVIDER_BROKERS = 'edge2ai-1.dim.local:9092'
+KAFKA_PROVIDER_BROKERS = '{}:9092'.format(get_hostname())
 KAFKA_PROVIDER_PROTOCOL = 'plaintext'
 
 SR_PROVIDER_NAME = 'sr'

setup/terraform/resources/setup.sh

@@ -229,15 +229,15 @@ EOF
   chown -R root:root /opt/cloudera/cem/${EFM_BASE_NAME}
   sed -i.bak 's#APP_EXT_LIB_DIR=.*#APP_EXT_LIB_DIR=/usr/share/java#' /opt/cloudera/cem/efm/conf/efm.conf
   sed -i.bak \
-'s#^efm.server.address=.*#efm.server.address=edge2ai-1.dim.local#;'\
+'s#^efm.server.address=.*#efm.server.address='"${LOCAL_HOSTNAME}"'#;'\
 's#^efm.server.port=.*#efm.server.port=10088#;'\
 's#^efm.security.user.certificate.enabled=.*#efm.security.user.certificate.enabled=false#;'\
 's#^efm.nifi.registry.enabled=.*#efm.nifi.registry.enabled=true#;'\
-'s#^efm.nifi.registry.url=.*#efm.nifi.registry.url=http://edge2ai-1.dim.local:18080#;'\
+'s#^efm.nifi.registry.url=.*#efm.nifi.registry.url=http://'"${LOCAL_HOSTNAME}"':18080#;'\
 's#^efm.nifi.registry.bucketName=.*#efm.nifi.registry.bucketName=IoT#;'\
 's#^efm.heartbeat.maxAgeToKeep=.*#efm.heartbeat.maxAgeToKeep=1h#;'\
 's#^efm.event.maxAgeToKeep.debug=.*#efm.event.maxAgeToKeep.debug=5m#;'\
-'s#^efm.db.url=.*#efm.db.url=jdbc:postgresql://edge2ai-1.dim.local:5432/efm#;'\
+'s#^efm.db.url=.*#efm.db.url=jdbc:postgresql://'"${LOCAL_HOSTNAME}"':5432/efm#;'\
 's#^efm.db.driverClass=.*#efm.db.driverClass=org.postgresql.Driver#;'\
 's#^efm.db.password=.*#efm.db.password='"${THE_PWD}"'#' /opt/cloudera/cem/efm/conf/efm.properties
   if [[ $ENABLE_TLS == yes ]]; then
@@ -251,7 +251,7 @@ EOF
 's#^efm.server.ssl.trustStoreType=.*#efm.server.ssl.trustStoreType=jks#;'\
 's#^efm.server.ssl.trustStorePassword=.*#efm.server.ssl.trustStorePassword='"$THE_PWD"'#;'\
 's#^efm.security.user.certificate.enabled=.*#efm.security.user.certificate.enabled=true#;'\
-'s#^efm.nifi.registry.url=.*#efm.nifi.registry.url=https://edge2ai-1.dim.local:18433#' /opt/cloudera/cem/efm/conf/efm.properties
+'s#^efm.nifi.registry.url=.*#efm.nifi.registry.url=https://'"${LOCAL_HOSTNAME}"':18433#' /opt/cloudera/cem/efm/conf/efm.properties
   fi
   echo -e "\nefm.encryption.password=${THE_PWD}${THE_PWD}" >> /opt/cloudera/cem/efm/conf/efm.properties
 
@@ -268,10 +268,11 @@ EOF
   chown -R root:root /opt/cloudera/cem/${MINIFITK_BASE_NAME}
   rm -f /opt/cloudera/cem/minifi/conf/bootstrap.conf
   if [[ $ENABLE_TLS == yes ]]; then
-    sed "s/THE_PWD/$THE_PWD/" $BASE_DIR/bootstrap.conf.tls > /opt/cloudera/cem/minifi/conf/bootstrap.conf
+    SOURCE_BOOTSTRAP_CONF=$BASE_DIR/bootstrap.conf.tls
   else
-    cp $BASE_DIR/bootstrap.conf /opt/cloudera/cem/minifi/conf/bootstrap.conf
+    SOURCE_BOOTSTRAP_CONF=$BASE_DIR/bootstrap.conf
   fi
+  sed "s/THE_PWD/$THE_PWD/;s/LOCAL_HOSTNAME/$LOCAL_HOSTNAME/" $SOURCE_BOOTSTRAP_CONF > /opt/cloudera/cem/minifi/conf/bootstrap.conf
   /opt/cloudera/cem/minifi/bin/minifi.sh install
 
   echo "-- Disable services here for packer images - will reenable later"
@@ -475,23 +476,10 @@ fi
 export CLUSTER_HOST=$PUBLIC_DNS
 export CDSW_DOMAIN=cdsw.${PUBLIC_IP}.nip.io
 
-echo "-- Load cluster metadata"
-if [[ -f $BASE_DIR/clusters_metadata.sh ]]; then
-  source $BASE_DIR/clusters_metadata.sh
-  PEER_CLUSTER_ID=$(( (CLUSTER_ID/2)*2 + (CLUSTER_ID+1)%2 ))
-  PEER_PUBLIC_DNS=$(echo "$CLUSTERS_PUBLIC_DNS" | awk -F, -v pos=$(( PEER_CLUSTER_ID + 1 )) '{print $pos}')
-  PEER_PUBLIC_DNS=${PEER_PUBLIC_DNS:-$PUBLIC_DNS}
-else
-  CLUSTER_ID=0
-  PEER_CLUSTER_ID=0
-  PEER_PUBLIC_DNS=$PUBLIC_DNS
-fi
-export CLUSTER_ID PEER_CLUSTER_ID PEER_PUBLIC_DNS
-
 echo "-- Set /etc/hosts - Public DNS must come first"
-sed -i.bak '/edge2ai-1.dim.local/ d' /etc/hosts
+sed -i.bak "/${LOCAL_HOSTNAME}/ d" /etc/hosts
 sed -i '/^::1/d' /etc/hosts
-echo "$PRIVATE_IP $PUBLIC_DNS $PRIVATE_DNS edge2ai-1.dim.local" >> /etc/hosts
+echo "$PRIVATE_IP $PUBLIC_DNS $PRIVATE_DNS $LOCAL_HOSTNAME" >> /etc/hosts
 
 echo "-- Configure networking"
 hostnamectl set-hostname ${CLUSTER_HOST}

setup/terraform/tf/deploy_cdp.tf

@@ -26,6 +26,7 @@ resource "null_resource" "deploy_cdp" {
       "set -o pipefail",
       "set -o xtrace",
       "trap 'echo Return code: $?' 0",
+      "sudo bash -c 'cat /tmp/resources/clusters_metadata.sh >> /etc/profile'",
       "# Prepare resources",
       "chmod +x /tmp/resources/*sh",
       "# Deploy workshop",

workshop_cdc.adoc

@@ -141,7 +141,7 @@ Let's start by connecting to PostgreSQL and creating the table.
 +
 [source,shell]
 ----
-psql --host edge2ai-1.dim.local --port 5432 --username cdc_user cdc_test
+psql --host localhost --port 5432 --username cdc_user cdc_test
 ----
 +
 When connected you should see the `psql` prompt, with the database name, as shown below:
@@ -231,7 +231,7 @@ Set the following required properties in your statement:
 [source,yaml]
 ----
 connector:     postgres-cdc
-hostname:      edge2ai-1.dim.local
+hostname:      <CLUSTER_HOSTNAME>
 username:      cdc_user
 password:      supersecret1
 database-name: cdc_test
@@ -259,7 +259,7 @@ CREATE TABLE transactions_cdc (
   name STRING
 ) WITH (
   'connector' = 'postgres-cdc',
-  'hostname' = 'edge2ai-1.dim.local',
+  'hostname' = '<CLUSTER_HOSTNAME>',
   'username' = 'cdc_user',
   'password' = 'supersecret1',
   'database-name' = 'cdc_test',
@@ -339,7 +339,7 @@ image:images/cdc/job-status-running.png[width=400]
 +
 [source,shell]
 ----
-psql --host edge2ai-1.dim.local --port 5432 --username cdc_user cdc_test
+psql --host localhost --port 5432 --username cdc_user cdc_test
 ----
 +
 In the `psql` prompt, execute the following commands to insert one more record in the `transactions` table and update the record with id `101`:
@@ -373,7 +373,7 @@ Nevertheless, you can replicate the data to any other database accessible via JD
 +
 [source,shell]
 ----
-psql --host edge2ai-1.dim.local --port 5432 --username cdc_user cdc_test
+psql --host localhost --port 5432 --username cdc_user cdc_test
 ----
 +
 [source,sql]
@@ -411,7 +411,7 @@ image::images/cdc/primary-key-clause.png[width=400]
 [source,yaml]
 ----
 connector:  jdbc
-url:        jdbc:postgresql://edge2ai-1.dim.local:5432/cdc_test
+url:        jdbc:postgresql://<CLUSTER_HOSTNAME>:5432/cdc_test
 table-name: trans_replica
 username:   cdc_user
 password:   supersecret1
@@ -472,7 +472,7 @@ image:images/cdc/kafka-json-template.png[width=300]
 [source,yaml]
 ----
 connector:                    kafka
-properties.bootstrap.servers: edge2ai-1.dim.local:9092
+properties.bootstrap.servers: <CLUSTER_HOSTNAME>:9092
 topic:                        trans_changelog
 key.format:                   json
 key.fields:                   id

workshop_dataviz.adoc

@@ -74,7 +74,7 @@ Connection name: Local Impala
 +
 [source,yaml]
 ----
-Hostname: <YOUR_CLUSTER_HOSTNAME> (something like: cdp.x.x.x.x.nip.io)
+Hostname: <CLUSTER_HOSTNAME> (something like: cdp.x.x.x.x.nip.io)
 Port #:   21050
 Username: [leave blank]
 Password: [leave blank]