entireX timeout

, ,
1

Hi,
How could i modify the time-out property for entireX?
I have tried to do it in this way: “System.setProperty(“entirex.timeout”, “10”)”, but it’s still not working.
Any suggestions?
Thks

2

When setting the timout value at the system level, the property is called

ETB_TIMEOUT
3

Could I use that property to terminate the wait time for free sockets?

4

Could you please describe what you are trying to achieve or what your problem is ?

5

  1. I had 2 brokers’ instances → Broker broker = new Broker DEFAULT_BROKERID, userId);

  2. I created threads and thrown 100 to eachone.

  3. The result was:

Caused by: Broker Error 0074 0074: WAIT timeout occurred
at com.softwareag.entirex.aci.d.c(Unknown Source)
at com.softwareag.entirex.aci.d.e(Unknown Source)
at com.softwareag.entirex.aci.d.a(Unknown Source)
at com.softwareag.entirex.aci.BrokerService.sendReceive(Unknown Source)
at com.softwareag.entirex.aci.BrokerService.sendReceive(Unknown Source)
at com.softwareag.entirex.aci.RPCService.a(Unknown Source)
at com.softwareag.entirex.aci.RPCService.a(Unknown Source)
at com.softwareag.entirex.aci.RPCService.callServer(Unknown Source)

How could i solve this problem?

Thks a lot

6

First of all, ETB_TIMEOUT denotes the transport timeout, i.e. how long the stub will wait for a response to a request from the opposite site.

For the 00740074 response, if I were you, I would take a look at the documentation, which says:

00740074 Wait Timeout Occurred 

Explanation The period specified by the WAIT field in the API has elapsed without any reply or request arriving to satisfy the Broker call. The conversation still exists and can continue. 
 
Action Check the following situations:

In a RECEIVE for servers, this is an information message when not waiting for a specific conversation. No request is available for processing. The RECEIVE is normally issued again. 

When this occurs for a specific conversation, a RECEIVE can be issued to wait again for the data/message, or the conversation can be aborted with an EOC call. 

When this occurs in a client in non-conversational mode, the call can be retried. 

Check if the partner is working correctly.
7

Ok,
How could i control this parameter?
Thnks for ur request
Kindest regards
Elena

8

Use the method setDefaultWaittime from Class BrokerService.

9

I tried this way but it isn’t working yet.
Could it have another wait-time in the transport?Also: timeout brokerservices and entireX
Kindest regards
Elena

10

before you go raising all the timeouts up, do you know why your partner applications are not responding within the existing time?

You indicate that you are starting 100’s of threads at a time, so apparently all of those hundreds of threads are attempting to make calls to a back-end service at the same time: is the back-end server able to service these requests? Do you have enough server instances running in the back end to process the requests? Does the server process have access to enough resources (database, memory, etc) to be able to run simultaneously?

11

Elena, is this a CLIENT or a SERVER ?

With 100 threads I guess it’s a serer, and for this it is perfectly valid to receive 00740074, this is just a ‘ping’ from the Broker, telling you ‘sorry, no work for you within the given time limit’ - look at the message description, just re-issue the RECEIVE.