@venkii    You need to login to your secured NiFi-Registry and make sure all your NiFi nodes have been authorized for both the following "Special Privileges": 1. "Read" for "Can Manage Buckets" 2. "Can proxy user requests"   Click on wrench icon in upper right corner to manage your users in NiFi-Registry. Then find your NiFi nodes in the list of USERS and click on the "manage user" pencil icon to the far right side.   Hope this helps, Matt ... View more

@venk  What you have run into at this point is a known issue. Your cluster was originally setup and running unsecured over HTTP port 8080. NiFi records the details of the nodes that are part of the cluster.  It does that so on later restarts it know that it should still be waiting on additional nodes to join before allowing users to make changes to the canvas. The downside to this is that when you switched to being secured over HTTPS on port 9091, the cluster now thinks you should have twice the number of nodes as there really are. But this is an easy fix.  Within your NiFi's conf directory you will find the file "state-management.xml". Inside that file you will find a section for NiFi's "local-provider" that will contain the directory where you can find your local state.  This path is normally the same on every node. Shutdown your NiFi and go to this directory on every node in your cluster and delete the contents within that state directory.  Restart your NiFi and it will only create new entries for your secured nodes. https://issues.apache.org/jira/browse/NIFI-7255 Hope this helps, Matt ... View more

@Koffi    When a NiFi node attempts to connect to an existing NiFi cluster, there are three files that are checked to make sure they match exactly between the connecting node and the existing copies in the cluster.  Those files are: 1. flow.xml.gz 2. users.xml (will only exist if NiFi is secured over https) 3. authorizations.xml (not to be confused with the NiFi authorizers.xml file.  Will only exist if NiFi is secured over https) The output in the nifi-app.log of the node should explain exactly what the mismatch was the first time it tried to connect to the cluster.   Hope this helps, Matt     ... View more

@domR    i) Do List processors w/ timestamp tracking store state locally? --- If you are running a standalone NiFi and not a NiFi cluster, all state will be stored locally on disk. --- If clustered, this depends on the list processor and how it is configured.  The ListFile processor can be configured to store state locally or remotely depending on your use case.  For example a ListFile is added to a NiFi cluster and every node is listing from a local path not shared across all nodes, you would want each node to store the listFile state locally since it would be unique per node and other nodes have no access to the directory one each node.  If your listFile is listing against a mounted directory that is mounted to every node in the cluster, the listFile should be configured for remote and configured t run on primary node only. --- Other list based processors all store state locally ONLY when it is a standalone NiFi.  Clustered NiFi installs will trigger store to be stored in zookeeper. ii) Does this state survive NiFi restarts? --- Yes, local state is stored on disk in NiFI's local state directory.  Cluster/Remote state is store in zookeeper.  State configurable is handled by the state-management.xml configuration file.   iii) If running on primary node only, would this mean when another primary node is chosen, the List processor would list any files it hasn't tracked (and re-ingress a large backlog of files if still there)? --- When a primary node change occurs, the primary node only processors on the previous primary node are asked to stop executing and the same processors on the newly elected primary node are asked to start.  On the new node, that processor will retrieve that last known state stored in zookeeper for that component before executing.  There is a small chance for some limited data duplication.  When old elected primary node processors are asked to stop that does not kill active threads. If the processor is in the middle of execution and does not complete (update cluster state in ZK) before newly elected primary node pulls cluster state when it starts to execute, some files may be listed again by newly elected node, but it will not list from beginning. iv) What out-of-box solutions can help to get around the issue of non-persisted non-distributed listing, or do we need custom auditing triggering individual listings? ---  NiFi does persist state through node restarts.   Note; You can right click on a processor that stores state and select "view state" to see what has been stored. You can also right click on a processor and select "view usage" to open the embedded documentation for that component. The embedded documentation will contain a "State Management:" section that will tell you if the component stores state and if that state is stored locally or cluster (ZK).   Hope this helps, Matt ... View more

@Alexandros    Going to ask the simple question first... There are FlowFiles traversing the processors in this newly instantiated flow from your template, correct? --- The next thought would then be around authorizations (assuming your NiFi is secured)  1. Is your user running the provenance query authorized to "view provenance" and "view the data" on the components?  If these are set on the process group containing these processor components and not set on the components themselves, the components will inherit the polices from the process group. 2. Is this a NiFi cluster?  If so, make sure your NiFi nodes are also authorized to "view provenance" and "view the data".  When you authenticate in to NiFi and run a provenance query, that query is replicated to all nodes in your cluster.  Those query results are then returned to the node on which the originating request was made.  If that node is not authorized to view data returned from other nodes, it will not display.   --- Then we need to make sure provenance it still working: While you are seeing provenance events displayed for your existing flow, are those returns recent?  If you monitor the contents of your provenance_repository, do you see timestamps updating on the <num>.prov files?  Need to make sure provenance has not stopped working for some reason.  Also make sure you are using the WriteAheadProvenanceRepository implementation (should be default in 1.11) and not the PersistentProvenanceRepository implementation (configured in nifi.properties file).   Hope this helps, Matt ... View more

@Gubbi  Depending on which processor is being used to create your FlowFile from you source linux directory, you will likely have an "absolute.path" FlowFile attribute created on the FlowFile. absolute.path =  /users/abc/ 20200312 /gtry/   You can pass that FlowFile to an UpdateAttribute processor which can use NiFi Expression Language (EL) to extract the date from that absolute path in to a new FlowFile attribute Add new property (property name becomes new FlowFile attribute): Property:          Value: pathDate          ${absolute.path:getDelimitedField('4','/')}   The resulting FlowFile will have a new attribute: pathDate = 20200312 Now you can use that FlowFile attribute later when writing to your target directory in S3. I assume you would use the putS3Object processor for this? If so, you can configure the "Object Key" property with the following: /Users/datastore/${pathDate}/${filename}   NiFi EL will replace ${pathDate} with "20200312" and ${filename} will be replaced with "gyyy.csv".   Hope this helps you, Matt ... View more

@anil35759    If you create a NiFi template that includes a NiFi processor that references a controller service, that controller service will be included in the generated template.  So if you import and instantiate that template on to the canvas of another NiFi, the controller service will be added as well and be associated with the processor instantiate from that same template. https://nifi.apache.org/docs/nifi-docs/html/user-guide.html#templates   So I am not clear on why you need to import the controller service separately via the toolkit.   Controller services are not mapped to processors.  The association happens from the processor side.   The processor is mapped to a specific controller service via the controller services assigned uuid. Keep in mind that anything you can do within the NiFi Ui can also be accomplished using NiFi rest-api endpoints.  Using your browsers developer tools to capture the requests as you make them through the NiFi UI is a great way to learn how to interact with NiFi rest-api endpoints.  The rest-api endpoints specific to your NiFi release version can be found under help within your install. Here are the Apache NiFi rest-api docs for the latest Apache release version: https://nifi.apache.org/docs/nifi-docs/rest-api/index.html Essentially what you need to do here is to update the processor configuration to reference the UUID of whichever controller service you want it to use. curl 'http://<nifi-hostname>:<nifi-port>/nifi-api/processors/b10fd083-0170-1000-0000-00007f7c905f' -X PUT --data-binary '{"component":{"id":"b10fd083-0170-1000-0000-00007f7c905f","name":"ExecuteSQL","config":{"concurrentlySchedulableTaskCount":"1","schedulingPeriod":"0 sec","executionNode":"ALL","penaltyDuration":"30 sec","yieldDuration":"1 sec","bulletinLevel":"WARN","schedulingStrategy":"TIMER_DRIVEN","comments":"","autoTerminatedRelationships":[],"properties":{"Database Connection Pooling Service":"e60cb24c-95c5-3a97-bcb9-9e537006317d"}},"state":"STOPPED"},"revision":{"clientId":"affea95c-0170-1000-988b-73bf756785b3","version":2},"disconnectedNodeAcknowledged":false}   So you'll notice from above example, I am updating a processor's (ExecuteSQL) configuration so that the Database Connection Pooling Service is mapped to the UUID of my DBCPConnectionPool controller service's uuid (e60cb24c-95c5-3a97-bcb9-9e537006317d).   Note:  The "clientId" string can be anything.    Hope this helps, Matt   ... View more

@sfishman  In the "Replacement value" configuration property just hold the <shift> key and hit enter instead of using "\n". Here is example of what you would see: This will create the line return you are looking for.   Hope this helps, Matt ... View more

@domR    I see no issues with using a publish and consume design for triggering follow on flows. That provide a slightly more robust setup than using the postHTTP to ListenHTTP example I provided since any NiFi node can consume the trigger file. When using Kafka you will want to make sure you have the same number of partitions as you have consumer in the same consumer group.  When you add a consumeKafka processor to the NiFi canvas it is really being added to every node in your cluster and is configured the same on every node. Let's assume you have a 3 node NiFi cluster, that means you have at a minimum 3 consumers in that consumer group so your Kafka topic should have 3 partitions; otherwise, you may see a lot of rebalancing happen.  To improve performance even more you can increase the concurrent task on your consumeKafka processor.  With a 3 node NiFi cluster and a consumeKafka configured with 2 concurrent tasks, you now have 6 total consumers in the consumer group; therefore, your Kafka topic should have 6 partitions.   If a NiFi node goes down Kafka will assign multiple partitions to the same consumer, so no need to worry about messages not being consumed. Hope this information was useful in finding a solution, Matt ... View more

@varun_rathinam    Can you please elaborate on " processor drop the file and join with new files"? And also "content_repository backup i limit"?  <-- Are you referring to the "nifi.content.repository.archive.max.retention.period" and "nifi.content.repository.archive.max.usage.percentage" configuration settings in the nifi.properties file? Also sharing a screenshot of your current MergeContent processor's configuration along with more details around your use case.  What result are you seeing now and what is the desired result? The MergeContent processor takes multiple FlowFiles from same NiFi node and merges the content of those FlowFiles based on the processor's configuration in to one or more new FlowFile's per node.  The processor cannot merge FlowFiles residing on different NiFi nodes in an NiFi cluster into one FlowFile.   FlowFiles from the inbound connection queue are allocated to bins based on the following configuration properties: Correlation Attribute Name  <-- (Optional) when used, only FlowFile with same value in the configured FlowFile attribute will be placed in same bin. Maximum Number of Entries  <-- Maximum number of FlowFiles that can be allocated to a single bin before a new bin is used. Maximum Group Size  <-- (Optional) Maximum cumulative size of the content that can be allocated to a bin   When a "bin" is eligible to be merged is controlled by these configuration properties: Minimum Number of Entries  <-- If at end of thread execution (after all FlowFiles from inbound connection have been allocated to one or more bins) the number of FlowFiles allocated to a bin meets this min and meets configured min group size, the FlowFiles in that bin will be merged. Minimum Group Size  <-- Same as above Max Bin Age  <-- A bin that has not reached or exceeded both above min values will merge once the bin has had FlowFiles in it for this amount of time  Maximum number of Bins  <-- If FlowFile have been allocated to every bin and another bin is needed, the oldest bin will be forced to merge to free a bin.   It is possible that one or both min values are never reached if a Max bin setting is reached first.  This means that because of max additional FlowFiles can not be allocated to that bin and the only setting that will force that bin to merge is "Max Bin Age" or you run out of free bins. As far as bin Max values, NiFi really does not care about content size as it streams the merged FlowFiles content in to a new FlowFile and does not hold that content in memory.  NiFi can experience Out OF Memory (OOM) conditions if the number of FlowFiles Max is set too high since all the attributes for every FlowFile currently allocated to bin(s) is held in heap memory.  NiFi's allocated heap memory is set in the nifi.properties configuration file. So a Max number of entries should be limited to 10000 (but this varies based on memory availability and number and size of attributes on your FlowFiles.  You can use multiple MergeContent processors in series (one after another) to merge multiple merged FlowFiles in to even larger merged FlowFiles if desired.   Hope this helps with understanding the MergeContent processor, Matt ... View more

@Umakanth    The API is exposed out of the box, it is not something you need to enable. Every action you take while performing actions within the UI makes a call to the NiFi rest-api. When learning how to use the rest-api calls, you may find using the developer tools in your browser helpful. Open the developer tools  while you are accessing your NiFi UI.  Then perform some action and you will see those requests being made by your browser to NiFi.  In the below example (Using Chrome browser developer tools), I opened NiFi's summary UI from the global menu in the upper right corner of the UI: You'll notice that several requests were made.  I can write click on any one of those request and select "Copy as cURL" to copy the full request to the system clipboard. I can then paste the request in a terminal window and see what the rest-api call returns.  You will notice that the curl command that is copied will have numerous additional headers (-H) that are not always necessary depending on the rest-api endpoint being used.  Example: curl 'http://<nifi-hostname>:<nifi-port>/nifi-api/flow/process-groups/root/status?recursive=true' Of course you will need to parse the rest-api returns yourself to extract in many cases the specific details/stats you want to monitor. Hope this helps, Matt ... View more