def start_cluster(self, cluster):
nn = vu.get_namenode(cluster)
run.format_namenode(nn)
run.start_hadoop_process(nn, 'namenode')
for snn in vu.get_secondarynamenodes(cluster):
run.start_hadoop_process(snn, 'secondarynamenode')
rm = vu.get_resourcemanager(cluster)
if rm:
run.start_yarn_process(rm, 'resourcemanager')
run.start_dn_nm_processes(utils.get_instances(cluster))
run.await_datanodes(cluster)
hs = vu.get_historyserver(cluster)
if hs:
run.start_historyserver(hs)
oo = vu.get_oozie(cluster)
if oo:
run.start_oozie_process(self.pctx, oo)
hiveserver = vu.get_hiveserver(cluster)
if hiveserver:
run.start_hiveserver_process(self.pctx, hiveserver)
self._set_cluster_info(cluster)
def start_cluster(self, cluster):
nn = vu.get_namenode(cluster)
run.format_namenode(nn)
run.start_hadoop_process(nn, 'namenode')
for snn in vu.get_secondarynamenodes(cluster):
run.start_hadoop_process(snn, 'secondarynamenode')
rm = vu.get_resourcemanager(cluster)
if rm:
run.start_yarn_process(rm, 'resourcemanager')
run.start_all_processes(utils.get_instances(cluster),
['datanode', 'nodemanager'])
run.await_datanodes(cluster)
hs = vu.get_historyserver(cluster)
if hs:
run.start_historyserver(hs)
oo = vu.get_oozie(cluster)
if oo:
run.start_oozie_process(self.pctx, oo)
hiveserver = vu.get_hiveserver(cluster)
if hiveserver:
run.start_hiveserver_process(self.pctx, hiveserver)
self._set_cluster_info(cluster)
def _set_cluster_info(self, cluster):
nn = vu.get_namenode(cluster)
rm = vu.get_resourcemanager(cluster)
hs = vu.get_historyserver(cluster)
oo = vu.get_oozie(cluster)
info = {}
if rm:
info['YARN'] = {
'Web UI': 'http://%s:%s' % (rm.management_ip, '8088'),
'ResourceManager': 'http://%s:%s' % (rm.management_ip, '8032')
}
if nn:
info['HDFS'] = {
'Web UI': 'http://%s:%s' % (nn.management_ip, '50070'),
'NameNode': 'hdfs://%s:%s' % (nn.hostname(), '9000')
}
if oo:
info['JobFlow'] = {
'Oozie': 'http://%s:%s' % (oo.management_ip, '11000')
}
if hs:
info['MapReduce JobHistory Server'] = {
'Web UI': 'http://%s:%s' % (hs.management_ip, '19888')
}
ctx = context.ctx()
conductor.cluster_update(ctx, cluster, {'info': info})
def start_cluster(self, cluster):
nn = vu.get_namenode(cluster)
run.format_namenode(nn)
run.start_hadoop_process(nn, "namenode")
for snn in vu.get_secondarynamenodes(cluster):
run.start_hadoop_process(snn, "secondarynamenode")
rm = vu.get_resourcemanager(cluster)
if rm:
run.start_yarn_process(rm, "resourcemanager")
for dn in vu.get_datanodes(cluster):
run.start_hadoop_process(dn, "datanode")
run.await_datanodes(cluster)
for nm in vu.get_nodemanagers(cluster):
run.start_yarn_process(nm, "nodemanager")
hs = vu.get_historyserver(cluster)
if hs:
run.start_historyserver(hs)
oo = vu.get_oozie(cluster)
if oo:
run.start_oozie_process(oo)
self._set_cluster_info(cluster)
def _set_cluster_info(self, cluster):
nn = vu.get_namenode(cluster)
rm = vu.get_resourcemanager(cluster)
hs = vu.get_historyserver(cluster)
oo = vu.get_oozie(cluster)
info = {}
if rm:
info['YARN'] = {
'Web UI': 'http://%s:%s' % (rm.management_ip, '8088'),
'ResourceManager': 'http://%s:%s' % (rm.management_ip, '8032')
}
if nn:
info['HDFS'] = {
'Web UI': 'http://%s:%s' % (nn.management_ip, '50070'),
'NameNode': 'hdfs://%s:%s' % (nn.hostname(), '9000')
}
if oo:
info['JobFlow'] = {
'Oozie': 'http://%s:%s' % (oo.management_ip, '11000')
}
if hs:
info['MapReduce JobHistory Server'] = {
'Web UI': 'http://%s:%s' % (hs.management_ip, '19888')
}
ctx = context.ctx()
conductor.cluster_update(ctx, cluster, {'info': info})
def _set_cluster_info(self, cluster):
nn = vu.get_namenode(cluster)
rm = vu.get_resourcemanager(cluster)
hs = vu.get_historyserver(cluster)
oo = vu.get_oozie(cluster)
info = {}
if rm:
info["YARN"] = {
"Web UI": "http://%s:%s" % (rm.management_ip, "8088"),
"ResourceManager": "http://%s:%s" % (rm.management_ip, "8032"),
}
if nn:
info["HDFS"] = {
"Web UI": "http://%s:%s" % (nn.management_ip, "50070"),
"NameNode": "hdfs://%s:%s" % (nn.hostname(), "9000"),
}
if oo:
info["JobFlow"] = {"Oozie": "http://%s:%s" % (oo.management_ip, "11000")}
if hs:
info["MapReduce JobHistory Server"] = {"Web UI": "http://%s:%s" % (hs.management_ip, "19888")}
ctx = context.ctx()
conductor.cluster_update(ctx, cluster, {"info": info})
def start_cluster(self, cluster):
nn = vu.get_namenode(cluster)
run.format_namenode(nn)
run.start_hadoop_process(nn, 'namenode')
for snn in vu.get_secondarynamenodes(cluster):
run.start_hadoop_process(snn, 'secondarynamenode')
rm = vu.get_resourcemanager(cluster)
run.start_yarn_process(rm, 'resourcemanager')
for dn in vu.get_datanodes(cluster):
run.start_hadoop_process(dn, 'datanode')
run.await_datanodes(cluster)
for nm in vu.get_nodemanagers(cluster):
run.start_yarn_process(nm, 'nodemanager')
hs = vu.get_historyserver(cluster)
if hs:
run.start_historyserver(hs)
oo = vu.get_oozie(cluster)
if oo:
run.start_oozie_process(oo)
self._set_cluster_info(cluster)
def _get_hadoop_configs(node_group):
cluster = node_group.cluster
nn_hostname = vu.get_instance_hostname(vu.get_namenode(cluster))
dirs = _get_hadoop_dirs(node_group)
confs = {
'Hadoop': {
'fs.defaultFS': 'hdfs://%s:9000' % nn_hostname
},
'HDFS': {
'dfs.namenode.name.dir': ','.join(dirs['hadoop_name_dirs']),
'dfs.namenode.data.dir': ','.join(dirs['hadoop_data_dirs']),
'dfs.hosts': '%s/dn-include' % HADOOP_CONF_DIR,
'dfs.hosts.exclude': '%s/dn-exclude' % HADOOP_CONF_DIR
}
}
res_hostname = vu.get_instance_hostname(vu.get_resourcemanager(cluster))
if res_hostname:
confs['YARN'] = {
'yarn.nodemanager.aux-services': 'mapreduce_shuffle',
'yarn.resourcemanager.hostname': '%s' % res_hostname,
'yarn.resourcemanager.nodes.include-path': '%s/nm-include' % (
HADOOP_CONF_DIR),
'yarn.resourcemanager.nodes.exclude-path': '%s/nm-exclude' % (
HADOOP_CONF_DIR)
}
confs['MapReduce'] = {
'mapreduce.framework.name': 'yarn'
}
hs_hostname = vu.get_instance_hostname(vu.get_historyserver(cluster))
if hs_hostname:
confs['MapReduce']['mapreduce.jobhistory.address'] = (
"%s:10020" % hs_hostname)
oozie = vu.get_oozie(cluster)
if oozie:
hadoop_cfg = {
'hadoop.proxyuser.hadoop.hosts': '*',
'hadoop.proxyuser.hadoop.groups': 'hadoop'
}
confs['Hadoop'].update(hadoop_cfg)
oozie_cfg = o_helper.get_oozie_required_xml_configs(HADOOP_CONF_DIR)
if c_helper.is_mysql_enabled(cluster):
oozie_cfg.update(o_helper.get_oozie_mysql_configs())
confs['JobFlow'] = oozie_cfg
if c_helper.get_config_value(c_helper.ENABLE_SWIFT.applicable_target,
c_helper.ENABLE_SWIFT.name, cluster):
swift_configs = {}
for config in swift.get_swift_configs():
swift_configs[config['name']] = config['value']
confs['Hadoop'].update(swift_configs)
if c_helper.is_data_locality_enabled(cluster):
confs['Hadoop'].update(th.TOPOLOGY_CONFIG)
confs['Hadoop'].update({"topology.script.file.name":
HADOOP_CONF_DIR + "/topology.sh"})
return confs, c_helper.get_env_configs()
def _get_hadoop_configs(pctx, instance):
cluster = instance.node_group.cluster
nn_hostname = vu.get_instance_hostname(vu.get_namenode(cluster))
dirs = _get_hadoop_dirs(instance)
confs = {
"Hadoop": {"fs.defaultFS": "hdfs://%s:9000" % nn_hostname},
"HDFS": {
"dfs.namenode.name.dir": ",".join(dirs["hadoop_name_dirs"]),
"dfs.datanode.data.dir": ",".join(dirs["hadoop_data_dirs"]),
"dfs.hosts": "%s/dn-include" % HADOOP_CONF_DIR,
"dfs.hosts.exclude": "%s/dn-exclude" % HADOOP_CONF_DIR,
},
}
res_hostname = vu.get_instance_hostname(vu.get_resourcemanager(cluster))
if res_hostname:
confs["YARN"] = {
"yarn.nodemanager.aux-services": "mapreduce_shuffle",
"yarn.resourcemanager.hostname": "%s" % res_hostname,
"yarn.resourcemanager.nodes.include-path": "%s/nm-include" % (HADOOP_CONF_DIR),
"yarn.resourcemanager.nodes.exclude-path": "%s/nm-exclude" % (HADOOP_CONF_DIR),
}
confs["MapReduce"] = {"mapreduce.framework.name": "yarn"}
hs_hostname = vu.get_instance_hostname(vu.get_historyserver(cluster))
if hs_hostname:
confs["MapReduce"]["mapreduce.jobhistory.address"] = "%s:10020" % hs_hostname
oozie = vu.get_oozie(cluster)
if oozie:
hadoop_cfg = {"hadoop.proxyuser.hadoop.hosts": "*", "hadoop.proxyuser.hadoop.groups": "hadoop"}
confs["Hadoop"].update(hadoop_cfg)
oozie_cfg = o_helper.get_oozie_required_xml_configs(HADOOP_CONF_DIR)
if c_helper.is_mysql_enabled(pctx, cluster):
oozie_cfg.update(o_helper.get_oozie_mysql_configs())
confs["JobFlow"] = oozie_cfg
if c_helper.is_swift_enabled(pctx, cluster):
swift_configs = {}
for config in swift.get_swift_configs():
swift_configs[config["name"]] = config["value"]
confs["Hadoop"].update(swift_configs)
if c_helper.is_data_locality_enabled(pctx, cluster):
confs["Hadoop"].update(th.TOPOLOGY_CONFIG)
confs["Hadoop"].update({"topology.script.file.name": HADOOP_CONF_DIR + "/topology.sh"})
hive_hostname = vu.get_instance_hostname(vu.get_hiveserver(cluster))
if hive_hostname:
hive_cfg = {
"hive.warehouse.subdir.inherit.perms": True,
"javax.jdo.option.ConnectionURL": "jdbc:derby:;databaseName=/opt/hive/metastore_db;create=true",
}
if c_helper.is_mysql_enabled(pctx, cluster):
hive_cfg.update(
{
"javax.jdo.option.ConnectionURL": "jdbc:mysql://%s/metastore" % hive_hostname,
"javax.jdo.option.ConnectionDriverName": "com.mysql.jdbc.Driver",
"javax.jdo.option.ConnectionUserName": "******",
"javax.jdo.option.ConnectionPassword": "******",
"datanucleus.autoCreateSchema": "false",
"datanucleus.fixedDatastore": "true",
"hive.metastore.uris": "thrift://%s:9083" % hive_hostname,
}
)
proxy_configs = cluster.cluster_configs.get("proxy_configs")
if proxy_configs and c_helper.is_swift_enabled(pctx, cluster):
key = key_manager.API().get(context.current(), proxy_configs["proxy_password"])
password = key.get_encoded()
hive_cfg.update(
{
swift.HADOOP_SWIFT_USERNAME: proxy_configs["proxy_username"],
swift.HADOOP_SWIFT_PASSWORD: password,
swift.HADOOP_SWIFT_TRUST_ID: proxy_configs["proxy_trust_id"],
swift.HADOOP_SWIFT_DOMAIN_NAME: CONF.proxy_user_domain_name,
}
)
confs["Hive"] = hive_cfg
return confs
def _get_hadoop_configs(pctx, node_group):
cluster = node_group.cluster
nn_hostname = vu.get_instance_hostname(vu.get_namenode(cluster))
dirs = _get_hadoop_dirs(node_group)
confs = {
'Hadoop': {
'fs.defaultFS': 'hdfs://%s:9000' % nn_hostname
},
'HDFS': {
'dfs.namenode.name.dir': ','.join(dirs['hadoop_name_dirs']),
'dfs.namenode.data.dir': ','.join(dirs['hadoop_data_dirs']),
'dfs.hosts': '%s/dn-include' % HADOOP_CONF_DIR,
'dfs.hosts.exclude': '%s/dn-exclude' % HADOOP_CONF_DIR
}
}
res_hostname = vu.get_instance_hostname(vu.get_resourcemanager(cluster))
if res_hostname:
confs['YARN'] = {
'yarn.nodemanager.aux-services': 'mapreduce_shuffle',
'yarn.resourcemanager.hostname': '%s' % res_hostname,
'yarn.resourcemanager.nodes.include-path': '%s/nm-include' % (
HADOOP_CONF_DIR),
'yarn.resourcemanager.nodes.exclude-path': '%s/nm-exclude' % (
HADOOP_CONF_DIR)
}
confs['MapReduce'] = {
'mapreduce.framework.name': 'yarn'
}
hs_hostname = vu.get_instance_hostname(vu.get_historyserver(cluster))
if hs_hostname:
confs['MapReduce']['mapreduce.jobhistory.address'] = (
"%s:10020" % hs_hostname)
oozie = vu.get_oozie(cluster)
if oozie:
hadoop_cfg = {
'hadoop.proxyuser.hadoop.hosts': '*',
'hadoop.proxyuser.hadoop.groups': 'hadoop'
}
confs['Hadoop'].update(hadoop_cfg)
oozie_cfg = o_helper.get_oozie_required_xml_configs(HADOOP_CONF_DIR)
if c_helper.is_mysql_enabled(pctx, cluster):
oozie_cfg.update(o_helper.get_oozie_mysql_configs())
confs['JobFlow'] = oozie_cfg
if c_helper.is_swift_enabled(pctx, cluster):
swift_configs = {}
for config in swift.get_swift_configs():
swift_configs[config['name']] = config['value']
confs['Hadoop'].update(swift_configs)
if c_helper.is_data_locality_enabled(pctx, cluster):
confs['Hadoop'].update(th.TOPOLOGY_CONFIG)
confs['Hadoop'].update({"topology.script.file.name":
HADOOP_CONF_DIR + "/topology.sh"})
hive_hostname = vu.get_instance_hostname(vu.get_hiveserver(cluster))
if hive_hostname:
hive_cfg = {
'hive.warehouse.subdir.inherit.perms': True,
'javax.jdo.option.ConnectionURL':
'jdbc:derby:;databaseName=/opt/hive/metastore_db;create=true'
}
if c_helper.is_mysql_enabled(pctx, cluster):
hive_cfg.update({
'javax.jdo.option.ConnectionURL':
'jdbc:mysql://%s/metastore' % hive_hostname,
'javax.jdo.option.ConnectionDriverName':
'com.mysql.jdbc.Driver',
'javax.jdo.option.ConnectionUserName': '******',
'javax.jdo.option.ConnectionPassword': '******',
'datanucleus.autoCreateSchema': 'false',
'datanucleus.fixedDatastore': 'true',
'hive.metastore.uris': 'thrift://%s:9083' % hive_hostname,
})
proxy_configs = cluster.cluster_configs.get('proxy_configs')
if proxy_configs and c_helper.is_swift_enabled(pctx, cluster):
hive_cfg.update({
swift.HADOOP_SWIFT_USERNAME: proxy_configs['proxy_username'],
swift.HADOOP_SWIFT_PASSWORD: proxy_configs['proxy_password'],
swift.HADOOP_SWIFT_TRUST_ID: proxy_configs['proxy_trust_id'],
swift.HADOOP_SWIFT_DOMAIN_NAME: CONF.proxy_user_domain_name
})
confs['Hive'] = hive_cfg
return confs
def start_historyserver(self, cluster):
hs = vu.get_historyserver(cluster)
if hs:
run.start_historyserver(hs)
def _get_hadoop_configs(pctx, instance):
cluster = instance.node_group.cluster
nn_hostname = vu.get_instance_hostname(vu.get_namenode(cluster))
dirs = _get_hadoop_dirs(instance)
confs = {
'Hadoop': {
'fs.defaultFS': 'hdfs://%s:9000' % nn_hostname
},
'HDFS': {
'dfs.namenode.name.dir': ','.join(dirs['hadoop_name_dirs']),
'dfs.datanode.data.dir': ','.join(dirs['hadoop_data_dirs']),
'dfs.hosts': '%s/dn-include' % HADOOP_CONF_DIR,
'dfs.hosts.exclude': '%s/dn-exclude' % HADOOP_CONF_DIR
}
}
res_hostname = vu.get_instance_hostname(vu.get_resourcemanager(cluster))
if res_hostname:
confs['YARN'] = {
'yarn.nodemanager.aux-services': 'mapreduce_shuffle',
'yarn.resourcemanager.hostname': '%s' % res_hostname,
'yarn.resourcemanager.nodes.include-path': '%s/nm-include' % (
HADOOP_CONF_DIR),
'yarn.resourcemanager.nodes.exclude-path': '%s/nm-exclude' % (
HADOOP_CONF_DIR)
}
confs['MapReduce'] = {
'mapreduce.framework.name': 'yarn'
}
hs_hostname = vu.get_instance_hostname(vu.get_historyserver(cluster))
if hs_hostname:
confs['MapReduce']['mapreduce.jobhistory.address'] = (
"%s:10020" % hs_hostname)
oozie = vu.get_oozie(cluster)
if oozie:
hadoop_cfg = {
'hadoop.proxyuser.hadoop.hosts': '*',
'hadoop.proxyuser.hadoop.groups': 'hadoop'
}
confs['Hadoop'].update(hadoop_cfg)
oozie_cfg = o_helper.get_oozie_required_xml_configs(HADOOP_CONF_DIR)
if c_helper.is_mysql_enabled(pctx, cluster):
oozie_cfg.update(o_helper.get_oozie_mysql_configs())
confs['JobFlow'] = oozie_cfg
if c_helper.is_swift_enabled(pctx, cluster):
swift_configs = {}
for config in swift.get_swift_configs():
swift_configs[config['name']] = config['value']
confs['Hadoop'].update(swift_configs)
if c_helper.is_data_locality_enabled(pctx, cluster):
confs['Hadoop'].update(th.TOPOLOGY_CONFIG)
confs['Hadoop'].update({"topology.script.file.name":
HADOOP_CONF_DIR + "/topology.sh"})
hive_hostname = vu.get_instance_hostname(vu.get_hiveserver(cluster))
if hive_hostname:
hive_cfg = {
'hive.warehouse.subdir.inherit.perms': True,
'javax.jdo.option.ConnectionURL':
'jdbc:derby:;databaseName=/opt/hive/metastore_db;create=true'
}
if c_helper.is_mysql_enabled(pctx, cluster):
hive_cfg.update({
'javax.jdo.option.ConnectionURL':
'jdbc:mysql://%s/metastore' % hive_hostname,
'javax.jdo.option.ConnectionDriverName':
'com.mysql.jdbc.Driver',
'javax.jdo.option.ConnectionUserName': '******',
'javax.jdo.option.ConnectionPassword': '******',
'datanucleus.autoCreateSchema': 'false',
'datanucleus.fixedDatastore': 'true',
'hive.metastore.uris': 'thrift://%s:9083' % hive_hostname,
})
proxy_configs = cluster.cluster_configs.get('proxy_configs')
if proxy_configs and c_helper.is_swift_enabled(pctx, cluster):
hive_cfg.update({
swift.HADOOP_SWIFT_USERNAME: proxy_configs['proxy_username'],
swift.HADOOP_SWIFT_PASSWORD: proxy_configs['proxy_password'],
swift.HADOOP_SWIFT_TRUST_ID: proxy_configs['proxy_trust_id'],
swift.HADOOP_SWIFT_DOMAIN_NAME: CONF.proxy_user_domain_name
})
confs['Hive'] = hive_cfg
return confs
def start_historyserver(self, cluster):
hs = vu.get_historyserver(cluster)
if hs:
run.start_historyserver(hs)
def _get_hadoop_configs(pctx, node_group):
cluster = node_group.cluster
nn_hostname = vu.get_instance_hostname(vu.get_namenode(cluster))
dirs = _get_hadoop_dirs(node_group)
confs = {
'Hadoop': {
'fs.defaultFS': 'hdfs://%s:9000' % nn_hostname
},
'HDFS': {
'dfs.namenode.name.dir': ','.join(dirs['hadoop_name_dirs']),
'dfs.namenode.data.dir': ','.join(dirs['hadoop_data_dirs']),
'dfs.hosts': '%s/dn-include' % HADOOP_CONF_DIR,
'dfs.hosts.exclude': '%s/dn-exclude' % HADOOP_CONF_DIR
}
}
res_hostname = vu.get_instance_hostname(vu.get_resourcemanager(cluster))
if res_hostname:
confs['YARN'] = {
'yarn.nodemanager.aux-services':
'mapreduce_shuffle',
'yarn.resourcemanager.hostname':
'%s' % res_hostname,
'yarn.resourcemanager.nodes.include-path':
'%s/nm-include' % (HADOOP_CONF_DIR),
'yarn.resourcemanager.nodes.exclude-path':
'%s/nm-exclude' % (HADOOP_CONF_DIR)
}
confs['MapReduce'] = {'mapreduce.framework.name': 'yarn'}
hs_hostname = vu.get_instance_hostname(vu.get_historyserver(cluster))
if hs_hostname:
confs['MapReduce']['mapreduce.jobhistory.address'] = ("%s:10020" %
hs_hostname)
oozie = vu.get_oozie(cluster)
if oozie:
hadoop_cfg = {
'hadoop.proxyuser.hadoop.hosts': '*',
'hadoop.proxyuser.hadoop.groups': 'hadoop'
}
confs['Hadoop'].update(hadoop_cfg)
oozie_cfg = o_helper.get_oozie_required_xml_configs(HADOOP_CONF_DIR)
if c_helper.is_mysql_enabled(pctx, cluster):
oozie_cfg.update(o_helper.get_oozie_mysql_configs())
confs['JobFlow'] = oozie_cfg
if c_helper.is_swift_enabled(pctx, cluster):
swift_configs = {}
for config in swift.get_swift_configs():
swift_configs[config['name']] = config['value']
confs['Hadoop'].update(swift_configs)
if c_helper.is_data_locality_enabled(pctx, cluster):
confs['Hadoop'].update(th.TOPOLOGY_CONFIG)
confs['Hadoop'].update(
{"topology.script.file.name": HADOOP_CONF_DIR + "/topology.sh"})
return confs