cancel
Showing results for 
Search instead for 
Did you mean: 

Long time query not receive response from server

olafo987
Node

Hello
Please, I need your help. I am loading a graph using apoc.periodic.iterate and apoc.load.jdbc. When the processing time exceeds 15 - 20 minutes, it appears as if the driver disconnected from the server and did not receive a response from the server when it finished the task. To run the query use
session.run (queryToCall, txConfig)
No error traces appear either in the logs of the cluster servers or in the log of the process that launches the query. Simply, my process does not receive the notification that the server has finished executing the sentence that has been passed to it.

Do you know what may be happening? Any recommendation on how to correct this error?

In advance, thanks for your help.

Juan Caros Pérez C

2 REPLIES 2

Joel
Ninja
Ninja

My first thought would be to ask for more details about the situation and what you have tried.

  • do you have verbose logging turned on, both on the neo4j server and at the client? (to ensure you see the event). The connection closure/timeout should be an event logged in both places, I'd imagine.
  • I guess you are calling from python? You are try/catching for any errors right?

At first blush it sounds a bit like a timeout, network interrupt/disconnect, but more likely a server blowup/crash (e.g. out of memory), any of these should leave obvious fingerprints in the logs at both ends though, unless the db literally did a kill -9 on itself (which I've never seen), but the client would still log an abrupt disconnect.

olafo987
Node

Thanks Jole a lot for your answer. First, I'm using Scala and the log level in both server and my load process is debug and in my code always write log at error level when catch an exception. I'm suspited for a network disconnect but I wanted ask in case there is any other possible cause that I have not contemplated.

Regards
Juan Carlos Pérez C