Administrative request queue freezes forever

We are building a scenario for master / master replication with two servers and one instance each..
The API is used to execute this operation. First time this was executed in LAN (2 M) and all the process has completed successful in 15 minutes.
Same process needs to be done in WAN (64K), and now (in this WAN) we got events as follows:
-     when executing ADD_MASTER_DATABASE, both instances put a lot of administrative request in its queues. Those request are never executed. It remains in status `AWAIT_CALLBACK` forever. We kept the process running for 10 hours with no results.
-     Then, once there is no activity on queues we do a “shutdown” and restart both instances. So the queues starts to work until be empty.
Servers are Windows 2000 Advanced Server, running Oracle 8i Enterprise Edition 8.1.6.
Thanks for any help

2 things.
1. there are several timeouts which will kick in eventually. Generally though they take 5-10 minutes.
2. If this is not bad code on your part, then its a function of your connection mechanism which could be due to yoru OS, or whatever is in your connection path. For instance, I connect to my rmi server, and if I pull the cable, my RMI client is aware the instant I try to make another call. Or if there is an active call it is instantly interrupted. And exception is immediately thrown in either case. However, if I am tunnelling over ssh, through an ssh client, and I pull the cable, the java program does not know instantly because my ssh client I guess does not instantly reveal this information.

Similar Messages

  • con:async-request-queue

    Hi,
              I've created the jms-queue, which was described in the manifest.xml, between the <con:async-request-queue> and <con:async-request-error-queue> -tags.
              I've a problem with the async-request-queue... When I send a message, the message is treated right, but it stays on this queue. Normally when a message is send, it gets the message from the specific queue... I really don't get what's wrong...
              Has anyone an idea?
              Thx in advance
              Ann

    hi,
    thanks for the reply.
    Just tell me if my understanding is right. Let us consider the application server is getting the requests. When we Enable Request Flow Control (by selecting the checkbox), we are making the server to put the requests to a queue(if all the threads of the process is busy, it puts it to the queue.). By, low water mark and high water mark, we mention the minimum and maximum amount of requests to be put into a queue.
    If i am right, i am getting the following doubt. You have mentioned that this wont work in KXS (that is... appserver will accept any amount of requests). What if the requests exceed the Request Queue High Water Mark? Will the server be continuing to get the requests? Or will it hang?
    regards,
    desigan

  • Propagate JMS Request Queue to JMS response in Topic

    Hi guys
    I trying to create a composite application that give me the response for one request queue and receive the response in a topic JMS, but I cant propagate the received response to the composite application. I have the response in the mediator but the BPEL process are waiting for the response.
    I working in SOA Suite 11.1.1.2 and WLS 10.2
    My composite have a main BPEL process and 2 mediators with Adapters to JMS Queue and JMS Topic
    Someone know what I´m doing wrong???
    Thanks.

    Hi,
    The problem as described seems like it is specific to BPEL -- you might want to try posting to the BPEL forum at BPEL
    Tom

  • How could increase the number of waiting request queue?

    how could increase the number of waiting request queue?
    thanks,

    I assume you're asking because you're getting connection refused exceptions? Set the AcceptBacklog parameter in the config.xml/via the console. Here's a description of the parameter:
    http://e-docs.bea.com/wls/docs81/config_xml/Server.html#AcceptBacklog

  • New to JMS: How to have to Request Queue

    hi
    I have a problem.
    I have integrated Weblogic 10 with Webspere 7.
    I am able to communicate between the same using a binding file.
    Now, I have one Request Queue in weblogic which in turn connects to the MDB.
    how can i have two request Queue.
    In my case : I have two swift System which send request to my MDB. when it is only one Incoming Queue,then it is fine.
    Can messaging Topic help? any dummies document on these.
    please help
    Thanks in advances
    Suhel

    I am not sure I fully understand your application configuration and what you are trying to achieve.
    A topic is for an application logic where multiple consumers need to receive a copy of each message that is sent to a destination What. you want to achieve is multiple senders though.
    If you don't want the two swift systems to share the same Request Queue, you can configure another queue in WebLogic and deploy another MDB to listen to it. In other words, if you configure 2 request queues, you can deploy 2 MDBs, each listen on one queue. You can also use WebLogic cluster and distributed queues, where multiple physical queues share the same logic jnid name. Then you only need to deploy one MDB.
    If you give more details about your goal, you have a better chance to get helpful answers.
    Thanks,
    Dongbo

  • Dispather request queue and FIFO

    Are the request in the request queue processed on the basis of first in first out (FIFO)? thanks.

    Yes, The requests are processed on a first-in, first-out basis (FIFO).
    Manu

  • Administrative requests hang and default queue is gone

    Hi,
    We have a recurring problem with our WebLogic 6.1 SP 1 server.
    Periodically the server will hang and not respond to admin commands,
    not
    even a "java weblogic.Admin PING". It has to be stopped by killing
    the OS process.
    We've dumped the threads using SIGQUIT and examined them for
    deadlocks, but something more striking turned up: all the "default
    queue" threads were gone. In a healthy, live server there were
    something like 60 threads, several of them from the "default queue",
    but in our hung servers, these would be nonexistent.
    From what I understand, the default queue contains the threads that do
    most of the work, like servicing EJB requests, so it makes sense that
    if they're not there, the server would not respond to anything. But I
    don't see how they could die without there being any kind of
    information about their death. There's nothing about it in the
    weblogic server log, and nothing in stdout or stderr.
    Has anybody experienced the same behavior? Does anyone know what
    causes it?
    Thanks,
    Jim Doyle
    [email protected]

    This is surprising. Are you seeing any OutOfMemory errors or strain on the
    server's resources ?
    Is the server under load ?
    Jim Doyle wrote:
    Hi,
    We have a recurring problem with our WebLogic 6.1 SP 1 server.
    Periodically the server will hang and not respond to admin commands,
    not
    even a "java weblogic.Admin PING". It has to be stopped by killing
    the OS process.
    We've dumped the threads using SIGQUIT and examined them for
    deadlocks, but something more striking turned up: all the "default
    queue" threads were gone. In a healthy, live server there were
    something like 60 threads, several of them from the "default queue",
    but in our hung servers, these would be nonexistent.
    From what I understand, the default queue contains the threads that do
    most of the work, like servicing EJB requests, so it makes sense that
    if they're not there, the server would not respond to anything. But I
    don't see how they could die without there being any kind of
    information about their death. There's nothing about it in the
    weblogic server log, and nothing in stdout or stderr.
    Has anybody experienced the same behavior? Does anyone know what
    causes it?
    Thanks,
    Jim Doyle
    [email protected]

  • Administration of Queues (GUI / command line)

    hi *,
    since i am working with JMQ now for a few weeks one thing bothers me most about the software....
    i could not find out since now if there is any administration GUI / command line interface that enables me to do very typical things in one tool:
    create / delete queues
    create / delete messages
    look into messages & properties
    etc...
    is there any or am i missing the point? how do several customers deal with this issue?
    regards chris

    First, thanks to you both for spending your time writing a response.
    We really wouldn't need this in production, this is more for testing each step of our translations.
    We actually do most of our IQmanager management through stcmsctrlutil.exe, though I do know we were looking at the management console through JCAPS as maybe an alternate solution for working with OpenMQ (it turned out to not be the case, and didn't research it any further). We weren't using the JMS API for OpenMQ to test any of the JMS functionalities.
    We haven't really thought of dumping or teeing the messages to put them back in the system (though we have had to do that before in our old SRE system in production, it was a mess). What I'm doing through scripts is pulling a message off at each JMS destination (after every translation) and storing it in a folder. I do the same for our old system (we're migrating from SRE --> JCAPS 6) and comparing the XML to make sure the translations are identical. We're a pretty small team, so we don't use any products like LISA or any testing suites, just a bunch of homegrown tools that do exactly what we need.
    It sounds like the private API will give us the needed functionality.
    When we're not on such a tight deadline for regression testing, I'll sit down and re-do all of the regression tools to work with OpenMQ, even if I have to change everything to queues, develop it using more JCAPS components (for pulling messages off and dumping them to the file system), or even use the unsupported API. Until then, we'll just do our unit testing on IQManager and our load testing on OpenMQ.
    big thanks from my team at Yum,
    --Adam                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                               

  • Pulled network cable... RMI freezes forever??

    I have RMI basically doing some remote tasks on various servers and all goes well... unless I pull a network cable out of a machine that is an RMI Server.
    My client, at this point, issues a request to that "server" and sits... forever.
    What I really need is a way to know that the reason it is waiting is because the server is simply not in a ready state. If I time it out somehow, the timeout may not know if it's a long process... worth sticking around for... or if it's waiting for no reason at all. So, for example, if I have a process that takes a few minutes usually... I'd like it to hang out. But I'd like it to just give up if it can't even get a connection to the machine...
    What's really odd is that I'd expect some sort of exception to be thrown... but nada... it just hangs... and then actually answers the function call when I plug it back in.
    I've looked through these forums and all over the place to find an answer to this issue... and have had no luck.
    So, if anyone has any ideas... PLEASE help.
    Thanks in advance!!

    2 things.
    1. there are several timeouts which will kick in eventually. Generally though they take 5-10 minutes.
    2. If this is not bad code on your part, then its a function of your connection mechanism which could be due to yoru OS, or whatever is in your connection path. For instance, I connect to my rmi server, and if I pull the cable, my RMI client is aware the instant I try to make another call. Or if there is an active call it is instantly interrupted. And exception is immediately thrown in either case. However, if I am tunnelling over ssh, through an ssh client, and I pull the cable, the java program does not know instantly because my ssh client I guess does not instantly reveal this information.

  • Queue freezing won't deliver messages to mailboxes

    Every now and again my queue seems to build up and won't deliver until I stop and start mail services then flush the queue.
    Here is my log, It seems to start filling up right after it refreshed the postfix mail system. I then continually get the message that it was skipped, and will still be delivered. But the messages are not being delivered and people sending messages are getting bounce messages. I've pasted the bounce message below the log.
    Nov 16 10:24:34 acs postfix/smtpd[23179]: connect from nsmusict.flyers.stark.k12.oh.us[10.110.62.89]
    Nov 16 10:24:34 acs postfix/smtpd[23179]: 7D1ED760E8: client=nsmusict.flyers.stark.k12.oh.us[10.110.62.89]
    Nov 16 10:24:34 acs postfix/cleanup[23231]: 7D1ED760E8: message-id=<[email protected]>
    Nov 16 10:24:34 acs postfix/qmgr[23169]: 7D1ED760E8: from=<[email protected]>, size=4810, nrcpt=1 (queue active)
    Nov 16 10:24:51 acs postfix/smtpd[23171]: connect from unknown[10.110.47.165]
    Nov 16 10:24:52 acs postfix/smtpd[23171]: disconnect from unknown[10.110.47.165]
    Nov 16 10:24:57 acs postfix/postfix-script[23410]: refreshing the Postfix mail system
    Nov 16 10:24:57 acs postfix/master[896]: reload configuration /etc/postfix
    Nov 16 10:24:57 acs postfix/qmgr[23415]: 0E221760E2: skipped, still being delivered
    Nov 16 10:24:57 acs postfix/qmgr[23415]: 7D1ED760E8: skipped, still being delivered
    Nov 16 10:24:57 acs postfix/qmgr[23415]: B2559760DF: skipped, still being delivered
    Nov 16 10:25:34 acs postfix/smtpd[23179]: disconnect from nsmusict.flyers.stark.k12.oh.us[10.110.62.89]
    Nov 16 10:25:38 acs postfix/smtpd[23418]: connect from pkrm323t2.flyers.stark.k12.oh.us[10.110.50.78]
    Nov 16 10:25:38 acs postfix/smtpd[23418]: 4A46A760F0: client=pkrm323t2.flyers.stark.k12.oh.us[10.110.50.78]
    Nov 16 10:25:38 acs postfix/cleanup[23421]: 4A46A760F0: message-id=<[email protected]>
    Nov 16 10:25:38 acs postfix/qmgr[23415]: 4A46A760F0: from=<[email protected]>, size=827, nrcpt=1 (queue active)
    Nov 16 10:25:52 acs postfix/smtpd[23426]: connect from unknown[10.110.47.165]
    Nov 16 10:25:53 acs postfix/smtpd[23426]: disconnect from unknown[10.110.47.165]
    Nov 16 10:26:31 acs postfix/smtpd[23426]: connect from amsrm810t.flyers.stark.k12.oh.us[10.110.50.161]
    Nov 16 10:26:31 acs postfix/smtpd[23426]: lost connection after EHLO from amsrm810t.flyers.stark.k12.oh.us[10.110.50.161]
    Nov 16 10:26:31 acs postfix/smtpd[23426]: disconnect from amsrm810t.flyers.stark.k12.oh.us[10.110.50.161]
    Nov 16 10:26:38 acs postfix/smtpd[23426]: connect from chadmorrismacbook.flyers.stark.k12.oh.us[10.110.47.95]
    Nov 16 10:26:38 acs postfix/smtpd[23426]: 37453760F5: client=chadmorrismacbook.flyers.stark.k12.oh.us[10.110.47.95]
    Nov 16 10:26:38 acs postfix/cleanup[23421]: 37453760F5: message-id=<[email protected]>
    Nov 16 10:26:38 acs postfix/qmgr[23415]: 0E221760E2: skipped, still being delivered
    Nov 16 10:26:38 acs postfix/qmgr[23415]: 37453760F5: from=<[email protected]>, size=589, nrcpt=1 (queue active)
    Nov 16 10:26:38 acs postfix/qmgr[23415]: 7D1ED760E8: skipped, still being delivered
    Nov 16 10:26:38 acs postfix/qmgr[23415]: B2559760DF: skipped, still being delivered
    This is the mail system at host example.com.
    I'm sorry to have to inform you that your message could not
    be delivered to one or more recipients. It's attached below.
    For further assistance, please send mail to postmaster.
    If you do so, please include this problem report. You can
    delete your own text from the attached returned message.
                      The mail system
    <[email protected]>: Command time limit exceeded:
       "/usr/bin/cyrus/bin/deliver"
    Reporting-MTA: dns; example.com
    X-Postfix-Queue-ID: 37453760F5
    X-Postfix-Sender: rfc822; [email protected]
    Arrival-Date: Fri, 16 Nov 2007 10:26:38 -0500 (EST)
    Final-Recipient: rfc822; [email protected]
    Original-Recipient: rfc822;[email protected]
    Action: failed
    Status: 5.3.0
    Diagnostic-Code: x-unix; internal software error
    From: Chad Morris <[email protected]>
    Date: November 16, 2007 10:26:37 AM EST
    To: Jan Horning <[email protected]>
    Subject: test

    I'm thinking that this could happen soon after I send out a mass email(50-300 users at once) to one of my postfix email groups. It isn't consistent, because it could happen a few minutes after I send one of those mass mailings, but then it doesn't do it all sometimes either.....
    Logs aren't giving me anything other than it will say for the first message that won't deliver (queue active) like seen below. This is everything in the log from a message that would deliver to a message that wouldn't.
    Nov 19 11:34:30 acs postfix/pipe[47472]: C00AA790B0: to=<[email protected]>, relay=cyrus, delay=0.1, delays=0.03/0.01/0/0.05, dsn=2.0.0, status=sent (delivered via cyrus service)
    Nov 19 11:34:30 acs postfix/qmgr[46598]: C00AA790B0: removed
    Nov 19 11:34:39 acs postfix/smtpd[46694]: disconnect from amsrm807t.flyers.stark.k12.oh.us[10.110.50.184]
    Nov 19 11:34:49 acs postfix/smtpd[46788]: disconnect from barracuda.flyers.stark.k12.oh.us[10.110.40.125]
    Nov 19 11:34:53 acs postfix/smtpd[47901]: connect from unknown[10.110.47.165]
    Nov 19 11:34:54 acs postfix/smtpd[47901]: disconnect from unknown[10.110.47.165]
    Nov 19 11:35:01 acs postfix/smtpd[47907]: connect from acsmoodle.flyers.stark.k12.oh.us[10.110.40.113]
    Nov 19 11:35:01 acs postfix/smtpd[47907]: 8AC7E790B2: client=acsmoodle.flyers.stark.k12.oh.us[10.110.40.113]
    Nov 19 11:35:01 acs postfix/cleanup[47345]: 8AC7E790B2: message-id=<[email protected]2. oh.us>
    Nov 19 11:35:01 acs postfix/qmgr[46598]: 8AC7E790B2: from=<[email protected]>, size=835, nrcpt=1 (queue active)
    Nov 19 11:35:01 acs postfix/smtpd[47907]: disconnect from acsmoodle.flyers.stark.k12.oh.us[10.110.40.113]

  • Render queue taking FOREVER

    I'm trying to export a 12 second comp and it's taking an extremely long time.  I've tried and cancelled several times now, from both the AE render queue and bringing it into Media Encoder and exporting from there.  In both locations, no matter what settings I choose, it takes north of 10 minutes for anything to happen, and once it does show progress, the counter on time remaining generally starts at several hours and counts up from there, ranging anywhere from 5 to 50 hours remaining.
    I have no other programs open, and have shut down and restarted my computer twice with no effect.  The composition has three roto brush effects and two mocha tracks in it, not sure if that could be causing a delay but still seems like WAY too much time to me for a 12 second comp.  Also, the RAM preview only takes a minute or two to complete to where I can play smoothly start to finish, should that not be about the same as export time or am I off on that?
    Any help appreciated.  Working on a mid-2012 MacbookPro with a 2.3G Intel Core i7, 8G RAM and an Intel HD Graphics 4000 1G graphics card.

    Running Yosemite 10.10.1.  And yes, I've rendered successfully before.  Actually exported an earlier version of the same comp with one of the mocha tracks on it so my buddy I'm doing the work for could see it.  That one exported in no more than a couple minutes.  The only thing I can really see that's different is the rotoscoping, and the thing that seems to be time consuming is in the upper right "Info" window, there is a running count of "Roto Brush Propagating."  It counts up until it gets to the number of frames, then starts over with a different number.  Guessing it's doing this for every frame and maybe that's the time issue?
    Not sure if I'm using a Ray-traced renderer, so if that's something I'd have to set myself I'd guess not.

  • System Administration - Request Set

    I am getting a very wierd situation I created a request set which had 5 programs it was running fine but 2 days before I updated the one program and schedule this request set it doesn't work.
    Sometimes it pick the one program sometime two I don't know I did everything in my Test/Middle environment if somebody had a same problem or some clue please let me know.
    Thanks and regards,
    Zeeshan

    Zeeshan,
    How did you link those requests? What modification have you done? Please elaborate more so others can help.

  • CRM Administrator: Request to identify Resource API's

    Hi Team,
    We have a requirement so as to update the effective dates in the
    CRM Administrator > Resource Manager > Maintain Resources > Resources i.e, in the table 'JTF_RS_DEFRESOURCES_VL'.
    Can anyone please help me to identify if we have any API's to update the data.
    Many Thanks in advance.
    Best Regards
    Arun Kumar S.R
    Apps Associates

    Arun,
    Just want to remind you that updating the dates has a lot of implications. A lot of transactions that use CRM resources are date driven.
    The object you mentioned is not a base table. Otherwise the API suggested by Alka is good to use.
    Thanks
    Nagamohan

  • Pending requests in the Standard Manager queue

    We are currently facing a situation in one of our Oracle EBS R12 instances, the current number of Standard Manager processes is set to 20, cache size 20 and sleep time 15 . We frequently observe that the concurrent requests get queued up in the Standard Manager queue (Concurrent->Manager->Administer).We have observed that the request queue gets cleared out when we temporarily increase the Standard Manager process to 30/40.But, the issue occurs again after some time, when we reset Standard manager process to 20.
    Any help in this regard will be highly helpful.
    OS: OEL 5.5 (x86)
    EBS: 12.1.1
    DB: 11.1.0.7
    Install type: Multi-Node
    Regards,
    Rupdipt

    Please see the docs referenced in your other thread -- Pending requests in the Standard Manager queue
    Also, see old threads which discuss the same topic.
    https://forums.oracle.com/forums/search.jspa?threadID=&q=Concurrent+AND+Manager+AND+Performance&objID=c3&dateRange=all&userID=&numResults=15&rankBy=10001
    https://forums.oracle.com/forums/search.jspa?threadID=&q=Concurrent+AND+Manager+AND+Tuning&objID=c3&dateRange=all&userID=&numResults=15&rankBy=10001
    https://forums.oracle.com/forums/search.jspa?threadID=&q=Concurrent+AND+Manager+AND+Processes&objID=c3&dateRange=all&userID=&numResults=15&rankBy=10001
    https://forums.oracle.com/forums/search.jspa?threadID=&q=Concurrent+AND+Manager+AND+Pending&objID=c3&dateRange=all&userID=&numResults=15&rankBy=10001
    Thanks,
    Hussein

  • FBU Internal Job Queue Full for Synchronous Processing of Print Requests

    Hello All,
    We are getting a lot of errors in system log (SM21) with the
    Error : FBU Internal Job Queue Full for Synchronous Processing of Print Requests
    User :SAPSYS
    MNO:FBU
    =============================================================
    Documentation for system log message FB U :
    If a spool server is to process output requests in the order they were
    created using multiple work processes, the processing of requests for a
    particular device can only run on one work process. To do this an
    internal queue (limited size) is used.
    If too many requests are created for this device too quickly, the work
    process may get overloaded. This is recognized when, as in this case,
    the internal queue is exhausted.
    This can only be solved by reducing the print load or increasing
    processor performance or, if there is a connection problem to the host
    spooler, by improving the transfer of data to the host spooler.
    Increasing the number of spool work processes will not help, as
    requests for one device can only be processed by one work process. If
    processing in order of creation is not required, sequential request
    processing can be deactivated (second page of device configuration in
    Transaction SPAD). This allows several work processes to process
    requests from the same device thus alleviating the bottleneck.
    Enlarging the internal queue will only help if the overload is
    temporary. If the overload is constant, a larger queue will eventually
    also be overloaded.
    ===========================================================
    Can you please tell me how to proceed.
    Best Regards,
    Pratyusha

    Solution is here:
    412065 - Incorrect output sequence of output requests
    Reason and Prerequisites
    The following messages appear in the developer trace (dev_trc) of the SPO processes or in the syslog:
    S  *** ERROR => overflow of internal job queue [rspowunx.c   788]
    Syslog Message FBU:
    Internal job queue for synchronous request processing of output requests full
    The "request processing sequence compliance" on a spool server with several SPO processes only works provided the server-internal job queue (see Note 118057) does not overflow. The size of this request queue is prepared using the rspo/global_shm/job_list profile parameter. The default value is 50 requests. However, if more output requests arrive for the spool server than can be processed (and the internal request queue is full as a result), more SPO processes are used to process the requests (in parallel), and the output sequence of the requests is no longer guaranteed.
    Solution
    Increase the rspo/global_shm/job_list profile parameter to a much larger value. Unfortunately, the value actually required cannot be found by "trial and error" because this queue contains all the incoming output requests on a spool server, not just the "sequence compliant" requests. A practical lower limit for this value represents the maximum sequence-compliant output requests for the above generated output device. If, for example, 1000 documents that should be output in sequence are issued from an application program to an output device, the queue must be able to hold 1000 entries so that it does not overflow if the SPO process processes the requests at a maximum low-speed.

Maybe you are looking for