Github user mans2singh commented on the issue:

    https://github.com/apache/nifi/pull/502
  
    Hey @pvillard31 - Thanks for your comments and I will try to address to 
your questions below:
    
    1.  In order to run integration tests, we need to run apache ignite 1.6 
locally (or create the proper xml configuration file - just like under the 
src/test/resources directory).  
    2. There are two integration tests (and a suite of unit tests) with the 
module.  The integration tests validate the data the contents of the flow with 
the data in the ignite cache.
    3. If the config file is not provided the processor creates a ignite client 
which tries to discover the ignite cluster on the localhost.  This could be 
used in case a single Nifi node is collocated with Ignite Cache node. When I 
run the processor using Nifi application, I can see the entry count 
corresponding the flow files in ignite cache by using the ignitevisorcmd.sh 
application.  If you see any problems with it, please let me know and I will to 
reproduce the scenario locally.
    4. Regarding provenance, Ignite tries to shield the clients from the 
servers by using a discovery spi.  While we can find out all the remote or 
local nodes in the cluster using this spi, we do not know which nodes will host 
the entry.  Getting all the nodes in the cluster will also impact the 
performance if we do it for the provenance of each flow file.  So, I've changed 
provenance to provide cache name which is how the client should access the 
entries.  Let me know your thoughts on it.
    5. The logs for Ignite can be handled by changing the log levels and I will 
look into it.
    6. I've been able to start/stop the cache and don't see the exceptions 
you've mentioned. If you have specific steps to reproduce them, please let me 
know.
    
    @apiri - My thought was that this processor is a egress component for 
streaming large volume/velocity data to the ignite cache.  If you have any 
thought/recommendation for me, please let me know.
    
    Thanks for all your advice.


---
If your project is set up for it, you can reply to this email and have your
reply appear on GitHub as well. If your project does not have this feature
enabled and wishes so, or if the feature is enabled but not working, please
contact infrastructure at infrastruct...@apache.org or file a JIRA ticket
with INFRA.
---

Reply via email to