def get_topic_config(self, topic):
"""Get configuration information for specified topic.
:rtype : dict of configuration
"""
try:
config_data = load_json(
self.get(
"/config/topics/{topic}".format(topic=topic)
)[0]
)
except NoNodeError as e:
# Kafka version before 0.8.1 does not have "/config/topics/<topic_name>" path in ZK and
# if the topic exists, return default dict instead of raising an Exception.
# Ref: https://cwiki.apache.org/confluence/display/KAFKA/Kafka+data+structures+in+Zookeeper.
topics = self.get_topics(topic_name=topic, fetch_partition_state=False)
if len(topics) > 0:
_log.info("Configuration not available for topic {topic}.".format(topic=topic))
config_data = {"config": {}}
else:
_log.error(
"topic {topic} not found.".format(topic=topic)
)
raise e
return config_data
def _get_entity_config(self, entity_type, entity_name, entity_exists):
"""Get configuration information for specified broker.
:entity_type : "brokers" or "topics"
:entity_name : broker id or topic name
:entity_exists : fn(entity_name) -> bool to determine whether an entity
exists. used to determine whether to throw an exception
when a configuration cannot be found for the given entity_name
:rtype : dict of configuration
"""
assert entity_type in (
"brokers", "topics"), "Supported entities are brokers and topics"
try:
config_data = load_json(
self.get("/config/{entity_type}/{entity_name}".format(
entity_type=entity_type, entity_name=entity_name))[0])
except NoNodeError as e:
if entity_exists(entity_name):
_log.info(
"Configuration not available for {entity_type} {entity_name}."
.format(
entity_type=entity_type,
entity_name=entity_name,
))
config_data = {"config": {}}
else:
_log.error("{entity_type} {entity_name} not found".format(
entity_type=entity_type, entity_name=entity_name))
raise e
return config_data
def get_topic_config(self, topic):
"""Get configuration information for specified topic.
:rtype : dict of configuration
"""
try:
config_data = load_json(
self.get("/config/topics/{topic}".format(topic=topic))[0])
except NoNodeError as e:
# Kafka version before 0.8.1 does not have "/config/topics/<topic_name>" path in ZK and
# if the topic exists, return default dict instead of raising an Exception.
# Ref: https://cwiki.apache.org/confluence/display/KAFKA/Kafka+data+structures+in+Zookeeper.
topics = self.get_topics(topic_name=topic,
fetch_partition_state=False)
if len(topics) > 0:
_log.info(
"Configuration not available for topic {topic}.".format(
topic=topic))
config_data = {"config": {}}
else:
_log.error("topic {topic} not found.".format(topic=topic))
raise e
return config_data
def get_broker_metadata(self, broker_id):
try:
broker_json, _ = self.get(
"/brokers/ids/{b_id}".format(b_id=broker_id))
except NoNodeError:
_log.error("broker '{b_id}' not found.".format(b_id=broker_id), )
raise
return load_json(broker_json)
def get_pending_plan(self):
"""Read the currently running plan on reassign_partitions node."""
reassignment_path = '{admin}/{reassignment_node}'\
.format(admin=ADMIN_PATH, reassignment_node=REASSIGNMENT_NODE)
try:
result = self.get(reassignment_path)
return load_json(result[0])
except NoNodeError:
return {}
def get_pending_plan(self):
"""Read the currently running plan on reassign_partitions node."""
reassignment_path = '{admin}/{reassignment_node}'\
.format(admin=ADMIN_PATH, reassignment_node=REASSIGNMENT_NODE)
try:
result = self.get(reassignment_path)
return load_json(result[0])
except NoNodeError:
return {}
def _fetch_partition_state(self, topic_id, partition_id):
"""Fetch partition-state for given topic-partition."""
state_path = "/brokers/topics/{topic_id}/partitions/{p_id}/state"
try:
partition_json, _ = self.get(
state_path.format(topic_id=topic_id, p_id=partition_id),
)
return load_json(partition_json)
except NoNodeError:
return {} # The partition has no data
def get_broker_metadata(self, broker_id):
try:
broker_json, _ = self.get(
"/brokers/ids/{b_id}".format(b_id=broker_id)
)
except NoNodeError:
_log.error(
"broker '{b_id}' not found.".format(b_id=broker_id),
)
raise
return load_json(broker_json)
def get_topic_config(self, topic):
"""Get configuration information for specified topic.
:rtype : dict of configuration
"""
try:
config_data = load_json(
self.get("/config/topics/{topic}".format(topic=topic))[0])
except NoNodeError as e:
_log.error("topic {topic} not found.".format(topic=topic))
raise e
return config_data
def get_broker_metadata(self, broker_id):
try:
broker_json = load_json(
self.get("/brokers/ids/{b_id}".format(b_id=broker_id))[0])
if (broker_json['host'] is None):
pattern = '(?:[SSL|INTERNAL|PLAINTEXTSASL].*://)?(?P<host>[^:/ ]+).?(?P<port>[0-9]*).*'
result = re.search(pattern, broker_json['endpoints'][0])
broker_json['host'] = result.group('host')
except NoNodeError:
_log.error("broker '{b_id}' not found.".format(b_id=broker_id), )
raise
return broker_json
def execute_plan(self,
plan,
allow_rf_change=False,
allow_rf_mismatch=False):
"""Submit reassignment plan for execution."""
reassignment_path = '{admin}/{reassignment_node}'\
.format(admin=ADMIN_PATH, reassignment_node=REASSIGNMENT_NODE)
plan_json = dump_json(plan)
topic_names_from_proposed_plan = set()
for partition in plan['partitions']:
topic_names_from_proposed_plan.add(partition['topic'])
base_plan = self.get_cluster_plan(
topic_names=list(topic_names_from_proposed_plan))
if not validate_plan(plan,
base_plan,
allow_rf_change=allow_rf_change,
allow_rf_mismatch=allow_rf_mismatch):
_log.error(
'Given plan is invalid. Aborting new reassignment plan ... {plan}'
.format(plan=plan))
return False
# Send proposed-plan to zookeeper
try:
_log.info('Sending plan to Zookeeper...')
self.create(reassignment_path, plan_json, makepath=True)
_log.info(
'Re-assign partitions node in Zookeeper updated successfully '
'with {plan}'.format(plan=plan), )
return True
except NodeExistsError:
_log.warning('Previous plan in progress. Exiting..')
_log.warning(
'Aborting new reassignment plan... {plan}'.format(plan=plan))
in_progress_plan = load_json(self.get(reassignment_path)[0])
in_progress_partitions = [
'{topic}-{p_id}'.format(
topic=p_data['topic'],
p_id=str(p_data['partition']),
) for p_data in in_progress_plan['partitions']
]
_log.warning(
'{count} partition(s) reassignment currently in progress:-'.
format(count=len(in_progress_partitions)), )
_log.warning(
'{partitions}. In Progress reassignment plan...'.format(
partitions=', '.join(in_progress_partitions), ), )
return False
except Exception as e:
_log.error(
'Could not re-assign partitions {plan}. Error: {e}'.format(
plan=plan, e=e), )
return False
def execute_plan(self, plan, allow_rf_change=False):
"""Submit reassignment plan for execution."""
reassignment_path = '{admin}/{reassignment_node}'\
.format(admin=ADMIN_PATH, reassignment_node=REASSIGNMENT_NODE)
plan_json = dump_json(plan)
base_plan = self.get_cluster_plan()
if not validate_plan(plan, base_plan, allow_rf_change=allow_rf_change):
_log.error('Given plan is invalid. Aborting new reassignment plan ... {plan}'.format(plan=plan))
return False
# Send proposed-plan to zookeeper
try:
_log.info('Sending plan to Zookeeper...')
self.create(reassignment_path, plan_json, makepath=True)
_log.info(
'Re-assign partitions node in Zookeeper updated successfully '
'with {plan}'.format(plan=plan),
)
return True
except NodeExistsError:
_log.warning('Previous plan in progress. Exiting..')
_log.warning('Aborting new reassignment plan... {plan}'.format(plan=plan))
in_progress_plan = load_json(self.get(reassignment_path)[0])
in_progress_partitions = [
'{topic}-{p_id}'.format(
topic=p_data['topic'],
p_id=str(p_data['partition']),
)
for p_data in in_progress_plan['partitions']
]
_log.warning(
'{count} partition(s) reassignment currently in progress:-'
.format(count=len(in_progress_partitions)),
)
_log.warning(
'{partitions}. In Progress reassignment plan...'.format(
partitions=', '.join(in_progress_partitions),
),
)
return False
except Exception as e:
_log.error(
'Could not re-assign partitions {plan}. Error: {e}'
.format(plan=plan, e=e),
)
return False
def get_json(self, path, watch=None):
"""Reads the data of the specified node and converts it to json."""
data, _ = self.get(path, watch)
return load_json(data) if data else None
def get_group_offsets(self, group, topic=None):
"""Fetch group offsets for given topic and partition otherwise all topics
and partitions otherwise.
{
'topic':
{
'partition': offset-value,
...
...
}
}
"""
group_offsets = {}
try:
all_topics = self.get_my_subscribed_topics(group)
except NoNodeError:
# No offset information of given consumer-group
_log.warning(
"No topics subscribed to consumer-group {group}.".format(
group=group, ), )
return group_offsets
if topic:
if topic in all_topics:
topics = [topic]
else:
_log.error(
"Topic {topic} not found in topic list {topics} for consumer"
"-group {consumer_group}.".format(
topic=topic,
topics=', '.join(topic for topic in all_topics),
consumer_group=group,
), )
return group_offsets
else:
topics = all_topics
for topic in topics:
group_offsets[topic] = {}
try:
partitions = self.get_my_subscribed_partitions(group, topic)
except NoNodeError:
_log.warning(
"No partition offsets found for topic {topic}. "
"Continuing to next one...".format(topic=topic), )
continue
# Fetch offsets for each partition
for partition in partitions:
path = "/consumers/{group_id}/offsets/{topic}/{partition}".format(
group_id=group,
topic=topic,
partition=partition,
)
try:
# Get current offset
offset_json, _ = self.get(path)
group_offsets[topic][partition] = load_json(offset_json)
except NoNodeError:
_log.error("Path {path} not found".format(path=path))
raise
return group_offsets
def get_multiple_topics(
self,
topic_names=None,
names_only=False,
fetch_partition_state=True,
):
"""Get information on all the available topics.
Topic-data format with fetch_partition_state as False :-
topic_data = {
'version': 1,
'partitions': {
<p_id>: {
replicas: <broker-ids>
}
}
}
Topic-data format with fetch_partition_state as True:-
topic_data = {
'version': 1,
'ctime': <timestamp>,
'partitions': {
<p_id>:{
replicas: [<broker_id>, <broker_id>, ...],
isr: [<broker_id>, <broker_id>, ...],
controller_epoch: <val>,
leader_epoch: <val>,
version: 1,
leader: <broker-id>,
ctime: <timestamp>,
}
}
}
Note: By default we also fetch partition-state which results in
accessing the zookeeper twice. If just partition-replica information is
required fetch_partition_state should be set to False.
"""
try:
if not topic_names:
topic_names = self.get_children("/brokers/topics")
except NoNodeError:
_log.error("Cluster is empty.")
return {}
if names_only:
return topic_names
topics_data = {}
for topic_id in topic_names:
try:
topic_info = self.get(
"/brokers/topics/{id}".format(id=topic_id))
topic_data = load_json(topic_info[0])
topic_ctime = topic_info[1].ctime / 1000.0
topic_data['ctime'] = topic_ctime
except NoNodeError:
_log.info(
"topic '{topic}' not found.".format(topic=topic_id), )
return {}
# Prepare data for each partition
partitions_data = {}
for p_id, replicas in six.iteritems(topic_data['partitions']):
partitions_data[p_id] = {}
if fetch_partition_state:
# Fetch partition-state from zookeeper
partition_state = self._fetch_partition_state(
topic_id, p_id)
partitions_data[p_id] = load_json(partition_state[0])
partitions_data[p_id][
'ctime'] = partition_state[1].ctime / 1000.0
else:
# Fetch partition-info from zookeeper
partition_info = self._fetch_partition_info(topic_id, p_id)
partitions_data[p_id][
'ctime'] = partition_info.ctime / 1000.0
partitions_data[p_id]['replicas'] = replicas
topic_data['partitions'] = partitions_data
topics_data[topic_id] = topic_data
return topics_data
def get_topics(
self,
topic_name=None,
names_only=False,
fetch_partition_state=True,
):
"""Get information on all the available topics.
Topic-data format with fetch_partition_state as False :-
topic_data = {
'version': 1,
'partitions': {
<p_id>: {
replicas: <broker-ids>
}
}
}
Topic-data format with fetch_partition_state as True:-
topic_data = {
'version': 1,
'ctime': <timestamp>,
'partitions': {
<p_id>:{
replicas: [<broker_id>, <broker_id>, ...],
isr: [<broker_id>, <broker_id>, ...],
controller_epoch: <val>,
leader_epoch: <val>,
version: 1,
leader: <broker-id>,
ctime: <timestamp>,
}
}
}
Note: By default we also fetch partition-state which results in
accessing the zookeeper twice. If just partition-replica information is
required fetch_partition_state should be set to False.
"""
try:
topic_ids = [topic_name] if topic_name else self.get_children(
"/brokers/topics",
)
except NoNodeError:
_log.error(
"Cluster is empty."
)
return {}
if names_only:
return topic_ids
topics_data = {}
for topic_id in topic_ids:
try:
topic_info = self.get("/brokers/topics/{id}".format(id=topic_id))
topic_data = load_json(topic_info[0])
topic_ctime = topic_info[1].ctime / 1000.0
topic_data['ctime'] = topic_ctime
except NoNodeError:
_log.info(
"topic '{topic}' not found.".format(topic=topic_id),
)
return {}
# Prepare data for each partition
partitions_data = {}
for p_id, replicas in six.iteritems(topic_data['partitions']):
partitions_data[p_id] = {}
if fetch_partition_state:
# Fetch partition-state from zookeeper
partition_state = self._fetch_partition_state(topic_id, p_id)
partitions_data[p_id] = load_json(partition_state[0])
partitions_data[p_id]['ctime'] = partition_state[1].ctime / 1000.0
else:
# Fetch partition-info from zookeeper
partition_info = self._fetch_partition_info(topic_id, p_id)
partitions_data[p_id]['ctime'] = partition_info.ctime / 1000.0
partitions_data[p_id]['replicas'] = replicas
topic_data['partitions'] = partitions_data
topics_data[topic_id] = topic_data
return topics_data
def get_group_offsets(self, group, topic=None):
"""Fetch group offsets for given topic and partition otherwise all topics
and partitions otherwise.
{
'topic':
{
'partition': offset-value,
...
...
}
}
"""
group_offsets = {}
try:
all_topics = self.get_my_subscribed_topics(group)
except NoNodeError:
# No offset information of given consumer-group
_log.warning(
"No topics subscribed to consumer-group {group}.".format(
group=group,
),
)
return group_offsets
if topic:
if topic in all_topics:
topics = [topic]
else:
_log.error(
"Topic {topic} not found in topic list {topics} for consumer"
"-group {consumer_group}.".format(
topic=topic,
topics=', '.join(topic for topic in all_topics),
consumer_group=group,
),
)
return group_offsets
else:
topics = all_topics
for topic in topics:
group_offsets[topic] = {}
try:
partitions = self.get_my_subscribed_partitions(group, topic)
except NoNodeError:
_log.warning(
"No partition offsets found for topic {topic}. "
"Continuing to next one...".format(topic=topic),
)
continue
# Fetch offsets for each partition
for partition in partitions:
path = "/consumers/{group_id}/offsets/{topic}/{partition}".format(
group_id=group,
topic=topic,
partition=partition,
)
try:
# Get current offset
offset_json, _ = self.get(path)
group_offsets[topic][partition] = load_json(offset_json)
except NoNodeError:
_log.error("Path {path} not found".format(path=path))
raise
return group_offsets
def get_json(self, path, watch=None):
"""Reads the data of the specified node and converts it to json."""
data, _ = self.get(path, watch)
return load_json(data) if data else None