- Added ClusterWorkload message to retrieve workload information from a
cluster coordinator
- Use cluster workload to return queued flow file count to site-to-site
client so that it can calculate distribution of data transfer
This closes#1084.
Signed-off-by: Bryan Bende <bbende@apache.org>
NIFI-2566: Removed storage of cluster roles from heartbeats and NodeConnectionStatus; use LeaderElectionManager to determine roles instead
NIFI-2566: Updated Heartbeats so that if a node is out-of-sync with cluster topology, cluster coordinator will provide updated information back to the nodes
NIFI-2566: Fixed issue that prevented standalone instance from starting by creating a standalone-instance version of the Leader Election Manager. Also added Controller Service enabled/disabled state to fingerprint rather than attempting to update the state when joining the cluster, as the implementation was incorrect and the correct implementation will be a rather significant effort that doesn't have to happen for 1.0.0 release
This closes#866
Signed-off-by: jpercivall <joepercivall@yahoo.com>
- Updated site-to-site client and server to support clustered nifi instances
NIFI-2274:
- Ensuring we use the correct URI when updating a connection.
This closes#530
Added unit tests for DN extraction.
Corrected typo in Javadoc.
Switched server/client socket logic for certificate extraction -- when the local socket is in client/server mode, the peer is necessarily the inverse.
Fixed unit tests.
Moved lazy-loading authentication access out of isDebugEnabled() control branch.
This closes#622
- Enable HTTP(S) for Site-to-Site communication
- Support HTTP Proxy in the middle of local and remote NiFi
- Support BASIC and DIGEST auth with Proxy Server
- Provide 2-phase style commit same as existing socket version
- [WIP] Test with the latest cluster env (without NCM) hasn't tested yet
- Fixed Buffer handling issues at asyc http client POST
- Fixed JS error when applying Remote Process Group Port setting from UI
- Use compression setting from UI
- Removed already finished TODO comments
- Added additional buffer draining code after receiving EOF
- Added inspection and assert code to make sure Site-to-Site client has
written data fully to output
stream
- Changed default nifi.remote.input.secure from true to false
This closes#497.
- Refactoring to allow requests to be replicated from a node to other nodes
- Renaming cluster node connection/read timeout properties.
- Renaming NCM DN to Cluster Coordinator DN.
- Fixing default values in properties.
- Starting to fix Spring context to load correctly in standalone mode.
- Using the cluster protocol to handle connection failures instead of heartbeats.
- Ensured replicate call is returned from ControllerResource.
- Ensure the appropriate classloader when serializing templates.
- Handling when the flow contents are null.
- This closes#488
- Started refactoring heartbeating mechanism, using a new package: org.apache.nifi.cluster.coordination
- Added configuration options for ZooKeeper username & password for heartbeat management. Also addressed issue where nodes that were previously disconnected were asked to disconnect upon restart
- Ensure that ACL is set properly when creating heartbeat node. Removed unused ControllerStartupFailureMessage.java
- Changed ZooKeeper ACL's so that container nodes that would not be sensitive are wide open and removed the usage of username & password when communicating with ZooKeeper. This was done specifically because username/password combination is considered a 'testing' feature that should not be used in production and is not supported by Apache Curator
- Refactored CuratorHeartbeatMonitor into an abstract heartbeat monitor that is responsible for processing heartbeats and CuratorHeartbeatMonitor that is responsible for retrieving heartbeat information
- Refactored so that heartbeats are sent to Cluster Coordinator directly instead of to ZooKeeper. ZooKeeper is used to know which node is the cluster coordinator but heartbeats to the Cluster Coordinator provide additional information about the nodes.
- Started refactoring heartbeating mechanism, using a new package: org.apache.nifi.cluster.coordination
- Added configuration options for ZooKeeper username & password for heartbeat management. Also addressed issue where nodes that were previously disconnected were asked to disconnect upon restart
- Changed ZooKeeper ACL's so that container nodes that would not be sensitive are wide open and removed the usage of username & password when communicating with ZooKeeper. This was done specifically because username/password combination is considered a 'testing' feature that should not be used in production and is not supported by Apache Curator
NIFI-1727:
- Refactored logic for merging HTTP Requests that are federated across cluster
NIFI-1745:
- Refactoring how HTTP Requests are replicated to nodes
- Bug fixes and continuing to work on replication refactoring. Still need to handle cluster locking and revisions
- Begin work on RevisionManager
- Resolved some issues that resulted from rebase
- Fixed URIs to align with new URI's that will be used in 1.0.0
- This closes#413
Added configuration options for ZooKeeper username & password for heartbeat management. Also addressed issue where nodes that were previously disconnected were asked to disconnect upon restart
Ensure that ACL is set properly when creating heartbeat node. Removed unused ControllerStartupFailureMessage.java
Changed ZooKeeper ACL's so that container nodes that would not be sensitive are wide open and removed the usage of username & password when communicating with ZooKeeper. This was done specifically because username/password combination is considered a 'testing' feature that should not be used in production and is not supported by Apache Curator
Refactored CuratorHeartbeatMonitor into an abstract heartbeat monitor that is responsible for processing heartbeats and CuratorHeartbeatMonitor that is responsible for retrieving heartbeat information
Refactored so that heartbeats are sent to Cluster Coordinator directly instead of to ZooKeeper. ZooKeeper is used to know which node is the cluster coordinator but heartbeats to the Cluster Coordinator provide additional information about the nodes.
Code cleanup and incorporate comments from peer review
This closes#323
Fixed nifi-utils pom.xml comment about additional dependencies. (+5 squashed commits)
Squashed commits:
[965b766] NIFI-1753 Removed temporary work-around of duplicate certificate conversion util method and added nifi-security-utils as dependency of nifi-utils.
[cd35f9b] NIFI-1753 Replaced legacy X.509 certificate declarations with new declarations in SSLSocketChannel and EndpointConnectionPool.
Temporary work-around of duplicate certificate conversion util method because nifi-utils cannot depend on nifi-security-utils.
[6420897] NIFI-1753 Replaced legacy X.509 certificate declarations with new declarations in PostHTTP.
[b9868ef] NIFI-1753 Added convenience method for extracting DN from peer certificate chain in SSL socket (canonical implementation to reduce code duplication and references to legacy certificate implementations).
Refactored logic retrieving legacy X.509 certificates with reference to convenience method in NodeProtocolSenderImpl.
Replaced logic retrieving legacy X.509 certificates with reference to convenience method in SocketProtocolListener.
Cleaned up exception handling in SocketProtocolListener.
Replaced legacy X.509 certificate declarations with new declarations in HandleHttpRequest (needs manual test).
[e2d1c35] NIFI-1753 Added convenience methods for converting legacy X.509 certificates and abstract certificates to correct X.509 format.
Added unit tests for certificate manipulation.
Replaced logic retrieving legacy X.509 certificates with new logic in NodeProtocolSenderImpl.
Added bcpkix (Bouncy Castle PKI implementation) dependency to nifi-standard-processors pom.
This closes#346.
Signed-off-by: Andy LoPresto <alopresto@apache.org>