So I have been trying to figure out what could be the flow of events if I am
planning to use JMS to handle 
long running stored procedure. I have a stored procedure(SP1), as soon as I
call this stored procedure, it records the details of the request and
returns a key and a message immediately.

There’s another procedure(SP2) that is launched asynchronously which is
responsible for gathering data from tables and 
dump it into 4 different tables.

I can keep track of the status (RUNNING, COMPLETE, ERROR)of the above
request by querying another table.


Since the above process is going to take hours, I am planning to use JMS
here. 

So, what should be my plan of action here? Here’s what I was thinking :

I get the parameters to call this stored procedure via REST end point. 
As soon as I call the stored procedure, I keep (my JMS Producer) sending
some message in every few minutes to the JMS Queue. Wondering if this should
be the status (RUNNING, COMPLETE , ERROR etc) that I can get from querying
another table or something else?
JMS Consumer would keep checking the message content every few minutes and
as soon as it finds “COMPLETE”, I can query that asynchronous stored
procedure(SP2) which is responsible for gathering data from 4 different
tables. 

I hope while the above process is going on, if I decide to send a call to
the SP1, let’s say via POSTMAN client, it won’t get stuck since the first
one is already running? This is where I was thinking of taking advantage of
JMS and making best use of it but not sure if everything I am trying to do
above is correct or not.

Please let me know if the above flow looks good/make sense or is there
anything more efficient I could do to achieve my goal. Let me know if I
could answer any more questions. Thanks





--
Sent from: http://activemq.2283324.n4.nabble.com/ActiveMQ-Dev-f2368404.html

Reply via email to