Question

7
Replies
425
Views
Posted: 1 year agoUpdated: 1 month ago

Dataflow Real-Time node status is Unknown

We have built a Dataflow using Twitter Dataset but when running the data flow we are getting an error :

Cannot execute the data flow run, because there are currently no functional data flow nodes in the cluster for the
RealTime
service. Ensure that there is at least one data flow node in the cluster with status 'Normal' via the
Data Flow Services landing page

When checking in Data Flow Services landing page, we see the node is having a status of UNKNOWN instead of NORMAL.

The dropdown for the execute link is not showing up and clicking on the execute link shows error.
The stack trace (Real-Time_NodeError - Copy.txt) is attached.

When checking in Decision Data Store, we see the same node showing status as Normal, with stop/repair options too.(Screenshots attached).

Could anyone please let us know how to resolve this ?

Posted: 1 year ago

What version of Pega platform are you using? On Latest version, on the Decisioning Services Landing page, please click on the Data flow tab and then you should select Real Time from the drop down and see if there is a node added and is in Normal status.
Posted: 1 year ago

Hi raghu_s, 


We are using Pega 8.1 


And on navigating to the data flow tab as you have mentioned, we are seeing the status as Unknown and has some stack trace(please see attached).


Thanks for your reply :)
Posted: 7 months ago

Hi, We are using Pega 8.1 and having the same issues. But the status of our node is "NORMAL" for "Data flow" tab whereas on "RealTIME Data Grid" tab , the status of the node is "JOINING_FAILED".I think our campaign is getting failed due to this with the below error on log: Unable to query VBD:


ERROR: Unable to query VBD: There are no functional nodes for Real-time Data Grid service


Please suggest necessary steps to be taken.
Posted: 7 months ago

Hi,


You can try adding the node after decommissioning it and if it again fails to join. Get the logs when it's failing to join and attach over here.
Posted: 1 month ago

Stop and decommission all the nodes reflecting status other than "Normal". Once removed, add the respective nodes back.


Let us know if that doesn't fix the issue.


Thank you,

Pawan
Posted: 1 month ago

Hi 


Adding to the above comment,if the issue still persists even after decomissioning the node from the landing page, then one of the reasons could be because ,the maximum no of active threads  utilization ,used for processing by the DF object, that may have exceeded the limit for that node.


Also, please try pausing or stopping all the other DF's and see if the issue still persists.


A support request can be raised to check for the thread dumps for better analysis.


Thanks

Renukavalli
Posted: 1 month ago

Hello Souvit,


As per my understanding, for a RT node to work, you should have a ADM node in normal state, and for an ADM to work, you would need a DDS node to be in normal state.


Let us know if the issue doesn't resolve after ensuring this.


Thank you,

Pawan