Hello! I don't think it could have affected CQ initial query, since queries/cache operations are supposed to block until they are successful, but not fail silently.
I'm not sure about connectTimeout, please try it and see. The main factor is that you expect your transactions to finish successfully. Setting low TxTimeoutOnPartitionMapExchange will lead to aborting of all your transactions when PME happens. Why would you want that, as opposed to the default when transactions finish and PME happens? Regards, -- Ilya Kasnacheev ср, 20 янв. 2021 г. в 17:50, VeenaMithare <v.mith...@cmcmarkets.com>: > >>It's possible that you have more than one network/interface and some > combinations are causing problems. Please try to specify localHost property > on every node pointing to a current actual external IP address of the node. > > This does seem to help the points 1,2,3,4. > Do you think this could have affected this issue : > > > http://apache-ignite-users.70518.x6.nabble.com/2-8-1-Continuous-Query-Initial-query-not-returning-any-result-sometimes-td35167.html > > >>5. Yes, it will do exponential backout based on those values. > Can the connectTimeout be set up for client side as well ? Would you know > why the default value is 600,000. > > >>6. I'm not sure what is the optimal value, it depends on your env. > Thanks. My understanding is that if a partition map exchange is detected > while a transaction is going on , timeout the transaction . In this case > shouldnt the TxTimeoutOnPartitionMapExchange be kept as small as possible ? > What are the other factors to consider here. > > > > > > > -- > Sent from: http://apache-ignite-users.70518.x6.nabble.com/ >
