delta/python-packages/kafka-python.git/kafka/cluster.py, branch state_change_callback_no_lock_close

Maintain shadow cluster metadata for bootstrapping (#1753)

2019-03-22T04:03:42+00:00

Retry bootstrapping after backoff when necessary (#1736)

2019-03-14T16:39:28+00:00

The current client attempts to bootstrap once during initialization, but if it fails there is no second attempt and the client will be inoperable. This can happen, for example, if an entire cluster is down at the time a long-running client starts execution.

This commit attempts to fix this by removing the synchronous bootstrapping from `KafkaClient` init, and instead merges bootstrap metadata with the cluster metadata. The Java client uses a similar approach. This allows us to continue falling back to bootstrap data when necessary throughout the life of a long-running consumer or producer.

Fix #1670

Ignore MetadataResponses with empty broker list (#1506)

2018-05-26T15:43:25+00:00

Document methods that return None

2018-05-23T22:19:35+00:00

If a valid broker in the cluster has no partitions, it will return None rather than an empty set.

Similarly updated a few other methods.

use absolute imports everywhere (#1362)

2018-02-06T20:10:45+00:00

KAFKA-3949: Avoid race condition when subscription changes during rebalance (#1364)

2018-02-03T00:36:30+00:00

Add ClusterMetadata documentation

2017-03-07T20:24:55+00:00

Drop old brokers when rebuilding broker metadata (#1005)

2017-03-03T20:51:21+00:00

Metadata with_partitions() (#787)

2016-08-02T01:14:59+00:00

  add method ClusterMetadata.with_partitions
  also fixup ClusterMetadata __str__

Vendor six 1.10.0

2016-08-01T21:47:40+00:00