Mikhali, This is a good piece of information. I think the problem may be due to too many requests queuing up at the listener. Have you tried with multiple listeners? For example a process with 20 threads, 10 connecting to the port 1414 and 10 to port 1415.
Thanks, Arvind -----Original Message----- From: mikhail malamud [mailto:[EMAIL PROTECTED] Sent: Friday, July 04, 2003 3:40 AM To: [EMAIL PROTECTED] Subject: Re: MQ Java Client NT Service and MQJE011 error Vaic - I am currently experiencing the same problem and have an open PMR with IBM in regards to it. I get MQJE001 under slightly different circumstances though. Here is the set up I have. Main thread launches n number of threads (agents). Each thread/agent establishes its own connection to the queue manager and performs gets and puts. The error happens when n > 16 intermittenly and permanently if n > 18. Error does not happen if there is a delay between the thread starts. Delay should be > 50milli secs. Note that I do not have any problems lauching 100 process each starting less than 16 threads all connecting to the same queue manager. Let me know if you want more info. ----- Original Message ----- From: Vaic, <mailto:[EMAIL PROTECTED]> Miroslav (CONS FIN, Trask, CONTRACTOR) To: [EMAIL PROTECTED] <mailto:[EMAIL PROTECTED]> Sent: Thursday, July 03, 2003 3:47 PM Subject: MQ Java Client NT Service and MQJE011 error Good morning to everybody. We are looking for a solution to a tough problem: We prepared a program in Java using Java client MQ files (com.ibm.mq.jar). The program is working perfectly when run interactively, but we need it to run as an NT service on Windows (using SRVANY utility from Microsoft Resource Kit). Whenever we try to start it, we get the following error: MQJE001: An MQException occurred: Completion Code 2, Reason 2059 MQJE011: Socket connection attempt refused We start the program using SRVANY utility from Microsoft and we expected that there may be some problems related to the network access limits for LocalSystem account etc. So, we tried several different accounts including local user account and domain user account - no success. We know that there is no problem with the QMGR because when running the program from a command prompt, everything is perfectly all right. We have also tried several different QMGRs including one that is local to our client (but also through TCP which is crucial to our architecture). What is quite interesting, running NETSTAT shows that there is a connection bound for this case: D:\500763723\Desktop>netstat Aktivní připojení Místní adresa Proto Cizí adresa Stav cizí adresy TCP W0759CGCFGE:1132 s0003cgcfge.adcz.cz:1026 NAVÁZÁNO TCP W0759CGCFGE:1202 pgfil01.cz.gcf.capital.ge.com:netbios-ssn NAVÁZÁNO TCP W0759CGCFGE:1580 udebsb.cz.gcf.capital.ge.com:22 NAVÁZÁNO TCP W0759CGCFGE:1733 s0003cgcfge.adcz.cz:epmap TIME_WAIT TCP W0759CGCFGE:1734 s0003cgcfge.adcz.cz:1026 TIME_WAIT TCP W0759CGCFGE:1736 s0001cgcfge.adcz.cz:epmap TIME_WAIT TCP W0759CGCFGE:1737 s0001cgcfge.adcz.cz:1026 TIME_WAIT TCP W0759CGCFGE:1414 W0759CGCFGE.adcz.cz:1699 NAVÁZÁNO TCP W0759CGCFGE:1699 W0759CGCFGE.adcz.cz:1414 NAVÁZÁNO (comment: NAVÁZÁNO means BOUND in Czech) Please, does anyone solve similar problem or would IBM support know any solution ? The product we use is WebSphere MQ 5.3 CSD 03 (Sun Solaris & Windows 2000). Thank for any help. Miroslav Vaic Instructions for managing your mailing list subscription are provided in the Listserv General Users Guide available at http://www.lsoft.com Archive: http://vm.akh-wien.ac.at/MQSeries.archive
