| Commit message (Collapse) | Author | Age | Files | Lines | |
|---|---|---|---|---|---|
| * | Do not log topic-specific errors in full metadata fetch (#1980) | Dana Powers | 2019-12-29 | 1 | -0/+4 |
| | | |||||
| * | Use dedicated connection for group coordinator (#1822) | Dana Powers | 2019-06-19 | 1 | -22/+14 |
| | | | | This changes the coordinator_id to be a unique string, e.g., `coordinator-1`, so that it will get a dedicated connection. This won't eliminate lock contention because the client lock applies to all connections, but it should improve in-flight-request contention. | ||||
| * | Remove unused imports (#1808) | Jeff Widman | 2019-05-22 | 1 | -2/+2 |
| | | |||||
| * | Maintain shadow cluster metadata for bootstrapping (#1753) | Dana Powers | 2019-03-21 | 1 | -10/+11 |
| | | |||||
| * | Retry bootstrapping after backoff when necessary (#1736) | Dana Powers | 2019-03-14 | 1 | -1/+23 |
| | | | | | | | | 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) | Dana Powers | 2018-05-26 | 1 | -1/+2 |
| | | |||||
| * | Document methods that return None | Jeff Widman | 2018-05-23 | 1 | -0/+3 |
| | | | | | | 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) | Kevin Tindall | 2018-02-06 | 1 | -3/+3 |
| | | |||||
| * | KAFKA-3949: Avoid race condition when subscription changes during rebalance ↵ | Dana Powers | 2018-02-02 | 1 | -0/+7 |
| | | | | | (#1364) | ||||
| * | Add ClusterMetadata documentation | Dana Powers | 2017-03-07 | 1 | -0/+14 |
| | | |||||
| * | Drop old brokers when rebuilding broker metadata (#1005) | Dana Powers | 2017-03-03 | 1 | -3/+6 |
| | | |||||
| * | Metadata with_partitions() (#787) | Dana Powers | 2016-08-01 | 1 | -1/+20 |
| | | | | | add method ClusterMetadata.with_partitions also fixup ClusterMetadata __str__ | ||||
| * | Vendor six 1.10.0six | Dana Powers | 2016-08-01 | 1 | -1/+1 |
| | | |||||
| * | Use MetadataRequest v1 for 0.10+ api_versionmetadata_v1 | Dana Powers | 2016-07-16 | 1 | -6/+37 |
| | | |||||
| * | Add rack to BrokerMetadata - it is always None when using MetadataRequest v0 | Dana Powers | 2016-07-16 | 1 | -8/+11 |
| | | |||||
| * | Avoid busy poll during metadata refresh failure with retry_backoff_ms (#733) | Dana Powers | 2016-06-20 | 1 | -0/+4 |
| | | |||||
| * | Remove unused importskafka-3318 | Dana Powers | 2016-04-05 | 1 | -1/+0 |
| | | |||||
| * | Update imports from kafka.common -> kafka.errors / kafka.structs | Dana Powers | 2016-04-05 | 1 | -2/+2 |
| | | |||||
| * | Improve kafka.cluster docstrings | Dana Powers | 2016-02-16 | 1 | -7/+65 |
| | | |||||
| * | Add lock to synchronize metadata updates and future handling | Dana Powers | 2016-02-15 | 1 | -18/+34 |
| | | |||||
| * | warn (not error) on LeaderNotAvailable - this is usually not cause for concern | Dana Powers | 2016-02-15 | 1 | -2/+2 |
| | | |||||
| * | Update ClusterMetadata comments | Dana Powers | 2016-02-15 | 1 | -7/+7 |
| | | |||||
| * | Track set of unauthorized_topics in ClusterMetadata | Dana Powers | 2016-02-15 | 1 | -0/+3 |
| | | |||||
| * | Remove unused cluster._version private attribute | Dana Powers | 2016-02-15 | 1 | -4/+1 |
| | | |||||
| * | Keep full PartitionMetadata from MetadataResponses | Dana Powers | 2016-02-15 | 1 | -6/+10 |
| | | |||||
| * | Implement KafkaConsumer.topics() | Dana Powers | 2016-02-01 | 1 | -2/+4 |
| | | | | | | - add ClusterMetadata.need_all_topic_metadata attribute - client requests metadata for all topics if attribute True | ||||
| * | Add available_partitions_for_topic() and partitions_for_broker() | Dana Powers | 2016-01-23 | 1 | -2/+23 |
| | | |||||
| * | Improve cluster.ttl() readability | Dana Powers | 2016-01-10 | 1 | -3/+7 |
| | | |||||
| * | Switch configs from attributes to dict to make passing / inspecting easier | Dana Powers | 2015-12-29 | 1 | -8/+12 |
| | | |||||
| * | Add ClusterMetadata.add_group_coordinator() | Dana Powers | 2015-12-28 | 1 | -0/+39 |
| | | |||||
| * | Rename Cluster -> ClusterMetadata; align with upstream Metadata class | Dana Powers | 2015-12-28 | 1 | -66/+121 |
| | | |||||
| * | Switch BrokerConnection to (mostly) non-blocking IO. | Dana Powers | 2015-12-17 | 1 | -1/+1 |
| | | | | | | | - return kafka.Future on send() - recv is now non-blocking call that completes futures when possible - update KafkaClient to block on future completion | ||||
| * | few small cleanups | Zack Dever | 2015-12-04 | 1 | -1/+1 |
| | | |||||
| * | Add simple Cluster class to manage broker metadata | Dana Powers | 2015-12-04 | 1 | -0/+91 |
 |
index : delta/python-packages/kafka-python.git | |
| github.com: mumrah/kafka-python.git |
| summaryrefslogtreecommitdiff |