Strange behavior in our receive locations:
RL_REPRESENTATIVE is waiting for a notification from the REPRESENTATIVE table (fields: (PK)id, fname, lname, etc).
RL_CLIENT_REPRESENTATIVE is waiting for a notification from the CLIENT_REPRESENTATIVE table (fields: (FK)id_rep, (FK)id_client).
When both of the locations are active and I commit a change in the CLIENT_REPRESENTATIVE.id_rep table I get a couple of warnings (apparently from RL_REPRESENTATIVE).
The adapter "WCF OracleDB" raised an error message. Details "System.InvalidOperationException: The notification query returned an error. Info="Error". Source="Data". Type="Change".
at Microsoft.Adapters.OracleDB.OracleDBInboundContract.Notification_TryReceive(OracleCommonExecutionHelper executionHelper, Message& wcfMessage)
at Microsoft.Adapters.OracleDB.OracleDBInboundContract.TryReceive(TimeSpan timeout, Message& message, IInboundReply& reply)
at Microsoft.ServiceModel.Channels.Common.Channels.AdapterInputChannel.TryReceive(TimeSpan timeout, Message& message)
at System.ServiceModel.Dispatcher.InputChannelBinder.TryReceive(TimeSpan timeout, RequestContext& requestContext)
at System.ServiceModel.Dispatcher.ErrorHandlingReceiver.TryReceive(TimeSpan timeout, RequestContext& requestContext)".
and
The adapter "WCF OracleDB" raised an error message. Details "The WCF service host at address oracledb://d01-isis:1521/D01ISIS/Dedicated?CallingTable=REPRESENTATIVE has faulted and as a result no more messages can be received on the corresponding receive location. To fix the issue, BizTalk Server will automatically attempt to restart the service host.".
Otherwise the process that is activated by a modification in CLIENT_REPRESENTATIVE takes place with no problems.
(If I update, instead, id_client in CLIENT_REPRESENTATIVE - the error comes from another receive location that subscribes to notifications from the CLIENT table.)
Two further clues:
If I disable RL_REPRESENTATIVE, the warnings won't appear.
If I update both CLIENT_REPRESENTATIVE.id_rep and REPRESENTATIVE.fname and commit both in the same transaction, the warnings won't appear.
Note that there are no triggers in either tables and all my timeouts are set to almost 24 hours.
I suspect that the FK constraint does its job in a way that ends up sending a notification to the port but I never get the actual message that I'm supposed to receive.
Question: Is there a parameter in Oracle that controls this behavior? Have any Biztalk devs ever run into this problem?
It could be:
the order in which things are saved.
a transactional problem, receive locations on seperate transactions, therefore it looks like related items are missing
a locking problem related to the transactional problem
The way to track it down would be to setup profiling on the database, and see exactly which commands are sent in which order.
In the end, the solution was to change a flag in all the receive locations (NotifyOnListenerStart) to "false".
After further development, this solution was not complete - the errant triggering is occurring again - so I'm toggling the check to off until I (or someone else) finds the correct solution.
Edit: This is a side-effect of the FK, changing its value seems to trigger a notification of a change in the table containing the PK (despite there not being any).
Related
I am trying to subscribe to iCloud changes but I keep getting an internal sever error. Saving and fethcing data from iCloud works, but I can not fetch nor save subscribtions.
Anyone know what could be the problem?
I have found the problem. CloudKit is very poorly documented when it comes to errors. Looks like you receive an internal server error when trying to subscribe for record changes, using a record type which does not exist yet. Weird, hope it helps someone in the future!
I have also noticed that once you encounter an error subsequent calls to CloudKit seem to fail silently. Perhaps its normal iOS behaviour, and I need to reset the error code if it gets set. Maybe someone else can comment.
I had the subscriptions working correctly until I added the subscriptionID. I then started receiving the "Internal server error". I was using the function:
let subscription = CKQuerySubscription(recordType: kActionFigureSpecificsRecord,
predicate: predicate,
subscriptionID: subscriptionId,
options: [.firesOnRecordCreation, .firesOnRecordUpdate, .firesOnRecordDeletion])
subscriptionID
The unique name of the subscription. This string must be unique in the specified database and must not be nil.
After reading the documentation, I realized that the subscriptionID needs to be unique, although the documentation wasn't particularly clear regarding unique relative to what. I believe the unique criteria is with respect to other database ID.names, such as other recordID.recordName. I had my subscriptionID string the same value as a recordID.recordName for the table on on which the subscription acted for the relevant record. When I appended _Subscription to the subscriptionID string (i.e., name), the subscription registered fine.
I am making an addon in Firefox, so I have a ChromeWorker - which is a privileged WebWorker. This is just a thread other then the mainthread.
In here I have no code but this (modified to make it look like not js-ctypes [which is the language for addons])
On startup I run this code, conn is a global variable:
conn = xcb_connect(null, null);
Then I run this in a 200ms interval:
evt = xcb_poll_for_event(conn);
console.log('evt:', evt);
if (!evt.isNull()) {
console.log('good got an event!!');
ostypes.API('free')(evt);
}
However evt is always null, I am never getting any events. My goal is to get all events on the system.
Anyone know what can cause something so simple to not work?
I have tried
xcb_change_window_attributes (conn, screens.data->root, XCB_CW_EVENT_MASK, values);
But this didn't fix it :(
The only way I can get it to work is by doing xcb_create_window xcb_map_window but then I get ONLY the events that happen in this created window.
You don't just magically get all events by opening a connection. There's only very few messages any client will receive, such as client messages, most others will only be sent to a client if it explicitly registered itself to receive them.
And yes, that means you have to register them on each and every window, which involves both crawling the tree and listening for windows being created, mapped, unmapped and destroyed and registering on them as well.
However, I would reconsider whether
My goal is to get all events on the system.
isn't an A-B problem. Why do you "need" all events? What do you actually want to do?
im facing a problem with my jms messeages.
Case is:
in my mainPS i give an employeeId on request which routes to BS to get employees with emploeeId >= than given. Response is employees collecion.
On response action in mainPS I loop through whole collection and id like to add to JMS queue one by one employee as simple message. In every loop iteration im seting user-defined transport header in if statement as msgName 'even' or 'odd' depends on employeeId on Outbound Request. On loged result i can see that headers have added properly.
Then (still for every employee) I publish my JMS-BS which adds message to queue.
When I want to pull my messages in JMSConsumerPS there are no any transport headers which i have added. Consumer has Get All Headers property set as 'YES'.
Im logging in consumer my whole
$inbound/ctx:transport
and what i get is:
<con:transport>
<con:uri>myURI</con:uri>
<con:mode>request</con:mode>
<con:qualityOfService>best-effort</con:qualityOfService>
<con:request xsi:type="jms:JmsRequestMetaData" xmlns:jms="http://www.bea.com/wli/sb/transports/jms" xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance">
<tran:headers xsi:type="jms:JmsRequestHeaders" xmlns:tran="http://www.bea.com/wli/sb/transports">
<jms:JMSDeliveryMode>2</jms:JMSDeliveryMode>
<jms:JMSExpiration>0</jms:JMSExpiration>
<jms:JMSMessageID>ID:<834866.1398327222060.0></jms:JMSMessageID>
<jms:JMSPriority>4</jms:JMSPriority>
<jms:JMSRedelivered>false</jms:JMSRedelivered>
<jms:JMSTimestamp>1398327222060</jms:JMSTimestamp>
<jms:JMSXDeliveryCount>1</jms:JMSXDeliveryCount>
</tran:headers>
<tran:encoding xmlns:tran="http://www.bea.com/wli/sb/transports">UTF-8</tran:encoding>
<jms:message-type>Text</jms:message-type>
</con:request>
Thanks for help.
Just a guess.
I was setting transport header in in-correct place. I had a "publish"
action in the proxy and I was setting transport header just before the
publish action. I moved the "transport header setting" to be done
inside the "publish" action/task . Now it works as expected.
Taken from https://community.oracle.com/thread/2155298?tstart=165, link most likely will be dead soon enough - Oracle community ...
Thank you very much for that. That was almost same issue. What I did was movig "transport header setting" into request action of published BS. Thing is i have tried it before asking but had no idea why it didnt work. Probably the reason could be that i turned on pass all headers through pipeline that time [?]. Had no idea if that can make such a mess. Thanks for replaying.
Here's what I'm attempting to do: Let's assume that you are in mail and create a New blank mail message, then enter some data into it, such as body copy, etc. (in my case, the message was created through scripting bridge using the "Mail Contents of this Page" from safari... the main purpose of this process for my application.)
From my application, I want to select that message and assign it to:
MailOutgoingMessage *myMessage;
so that I can programmatically add recipients. I've tried several ways of doing this which seemed logical, but so far I haven't found the right combination, and the header file doesn't seem to be very clear to me (I'm new to scripting bridge.)
My initial thought was to try this:
mailMessage = [[mail outgoingMessages] lastObject];
Which should grab the last outgoing message created. It seems to work in that I am able to add recipients to mailMessage (though there have been a few times that I received unexpected results when multiple outgoing messages exist, such as adding the recipients to the wrong message) but attempting to log the subject line of the message:
NSLog(#"Subject = %#",[mailMessage subject]);
always returns NULL even though there is a subject clearly viewable in the subject field of the message. NULL is returned for any other parameter as well.
I'm gathering it must be a problem with my assignment to mailMessage above, because the only time I receive a NULL for message properties (or receive unexpected results) is if I try to point mailMessage to an existing outgoing message. If I create the mail message with scripting bridge, then I can retrieve all of the properties correctly.
Does anyone understand the hierarchy of the Mail scripting enough to tell me why I am getting NULLs for the parameters using the above assignment for mailMessage? What would the simplest way be go grab my message so that I can add recipients and later call the:
[myMessage send];
method? Any insight would be helpful. I've spent a week going through the mail.h header file and am quite literally at a loss as to what else to try at this point.
There's no way to (send, get or set the properties of the outgoing message) that the user or Safari has created.
It's a bug (it stopped working since Mac OS X 10.4), or some privacy/security considerations.
Using MS Visual Studio 2008 C++ for Windows 32 (XP brand), I try to construct a POP3 client managed from a modeless dialog box.
Te first step is create a persistent object -say pop3- with all that Boost.asio stuff to do asynchronous connections, in the WM_INITDIALOG message of the dialog-box-procedure. Some like:
case WM_INITDIALOG:
return (iniPop3Dlg (hDlg, lParam));
Here we assume that iniPop3Dlg() create the pop3 heap object -say pointed out by pop3p-. Then connect with the remote server, and a session is initiated with the client’s id and password (USER and PASS commands). Here we assume that the server is in TRANSACTION state.
Then, in response to some user input, the dialog-box-procedure, call the appropriate function. Say:
case IDS_TOTAL: // get how many emails in the server
total (pop3p);
return FALSE;
case IDS_DETAIL: // get date, sender and subject for each email in the server
detail (pop3p);
return FALSE;
Note that total() uses the POP3’s STAT command to get how many emails in the server, while detail() uses two commands consecutively; first STAT to get the total and then a loop with the GET command to retrieve the content of each message.
As an aside: detail() and total() share the same subroutines -the STAT handle routine-, and when finished, both leaves the session as-is. That is, without closing the connection; the socket remains opened an the server in TRANSACTION state.
When any option is selected by the first time, the things run as expected, obtaining the desired results. But when making the second chance, the connection hangs.
A closer inspection show that the first time that the statement
socket_.get_io_service().run();
Is used, never ends.
Note that all asynchronous write and read routines uses the same io_service, and each routine uses socket_.get_io_service().reset() prior to any run()
Not also that all R/W operations also uses the same timer, who is reseted to zero wait after each operation is completed:
dTimer_.expires_from_now (boost::posix_time::seconds(0));
I suspect that the problem is in the io_service or in the timer, and the fact that subsequent executions occurs in a different load of the routine.
As a first approach to my problem, I hope that someone would bring some light in it, prior to a more detailed exposition of the -very few and simple- routines involved.
Have you looked at the asio examples and studied them? There are several asynchronous examples that should help you understand the basic control flow. Pay particular importance to the main event loop started by invoking io_service::run, it's important to understand control is not expected to return to the caller until the io_service has no more remaining work to do.