Strange QuickFix engine response to execution report message - quickfix

I am currently experiencing a strange problem with the open-source QuickFix engine. Our broker is sending some test trades with MsgType = 8 (Execution Report), and our QuickFix engine immediately replies with an exception saying "Unsupported Message Type". All of the tags in the broker's message appear to be legitimate.
Why is this happening and how can I resolve the issue?

I can't read minds, but it sounds like you may have implemented your application as a MessageCracker but forgotten to override the appropriate onMessage function. Note there is a separate onMessage overload for each FIX version of a message type, e.g. there are:
onMessage (const FIX40::ExecutionReport&, const FIX::SessionID&)
onMessage (const FIX41::ExecutionReport&, const FIX::SessionID&)
onMessage (const FIX42::ExecutionReport&, const FIX::SessionID&)
onMessage (const FIX43::ExecutionReport&, const FIX::SessionID&)
onMessage (const FIX44::ExecutionReport&, const FIX::SessionID&)
The default implementation of all of these methods throws an UnsupportedMessageType exception, which sounds like what you are seeing.

Related

How to get message body from message intercepted by ActiveMQServerMessagePlugin afterSend method

I intercept the message by public void afterSend(ServerSession session, Transaction tx, Message message, boolean direct, boolean noAutoCreateQueue, RoutingStatus result) throws ActiveMQException method. How do I get the message body? I try to get body by ((CoreMessage) message).getBodyBuffer().readUTF() but get Method threw 'java.lang.IndexOutOfBoundsException' exception. And in debugger body = null, but sended message has String body.
The first thing to note is that org.apache.activemq.artemis.api.core.Message#getBodyBuffer is deprecated and shouldn't be used.
If your message was sent using the core JMS client and it was sent as a javax.jms.TextMessage then you can use org.apache.activemq.artemis.api.core.Message#getStringBody to get the underlying byte array decoded as a String.

Flutter/Dart + gRPC Stream error: Stream was terminated by peer (errorCode: 2)

I'm having this intermittent error come up when making a call from a Dart gRPC client to a Go gRPC server. The exact error message is:
gRPC Error (code: 2, codeName: UNKNOWN, message: HTTP/2 error: Stream
error: Stream was terminated by peer (errorCode: 2)
This seems to occur frequently on creating the client stub with the channel and making a call to the server. Subsequent calls seems to work fine with a response returned correctly to the client. Then this error intermittently pops up again every n-th call (could be after 5 calls, 10 calls sometimes no errors until the 20th call).
When I use a different client (eg: BloomRPC or Java app I wrote), this does not happen no matter how hard or fast I spam the calls to the server. This kind of implies the server is probably not the issue.
Can anyone shine a light on what might be happening? Or provide some hints as to where to start looking/debugging? I've tried stepping through the code, but not really able to follow what is happening when the gRPC call is made/returned.
Steps in the client code:
create ClientChannel with ip, ports
final cc = ClientChannel(serverIP,
port: serverPort,
options:
const ChannelOptions(credentials: ChannelCredentials.insecure()
));
created a stub based on above channel
_userQueryStub = UserQueryGRPCClient(cc,
options: CallOptions(timeout: Duration(seconds: 30)));
call function from stub (repeatedly)
GetUserInformationRequest request = GetUserInformationRequest(userId: userId);
GetUserInformationResponse response = await _userQueryStub.getUserInformation(request);
Any help will be greatly appreciated!

How stop execution and send ack/nack with error message in mirth 3.4?

I'm new in mirth so sorry if my question may seems naive.
I've a mirth channel that recives hl7 messages, and this is fine, also I've some filters and transformers both in Source and Destination.
When all is fine at the end of destination I send an ACK with a message, for this for this purpose I've made this function in code Templates:
function getAck(success, detailMessage, statusMessage) {
if (!detailMessage)
detailMessage = success ? "Operation completed successfully" : "Some error occours";
if(!statusMessage)
statusMessage = detailMessage;
if (success) {
ack = ACKGenerator.generateAckResponse(connectorMessage.getRawData(), "AA", detailMessage);
resp = new Response(com.mirth.connect.userutil.Status.SENT, ack, statusMessage);
} else {
ack = ACKGenerator.generateAckResponse(connectorMessage.getRawData(), "AE", detailMessage);
resp = new Response(com.mirth.connect.userutil.Status.ERROR, ack, statusMessage, detailMessage);
}
return resp;
}
So I use ACKGenerator.generateAckResponse for creating an Ack and Response for send response at client. This work but only in destination and that's my problem.
If I get an error before destination (e.g. in filters, transformer, ...) I don't be able to stop execution and send an NACK with an explaination of the error and this is what I would like to do.
Am I wrong doing things in this way?
You can store a Response in the responseMap in any filters or transformers. Once you define a key in the responseMap, it should be available as a selection in the response drop down on the source tab of your channel (instead of picking a destination.)
Your current connector should stop processing a message with an ERROR status if you manually throw an exception after setting the desired value in the responseMap. If you are in a filter, you could also filter the message instead of throwing the exception.
If you are worried about an uncaught exception, you could initialize the responseMap variable with an "Unknown Error" message at the first point in the channel where your custom code is defined that affects messages directly (likely the source filter from your description, but could possibly be the pre-processor or attachment handler if you use those.) The expectation is that this would be replaced with a more descriptive error or a success if the message makes it all the way through to the end, but the channel will always have something to return.
There are filter and transformer in the "Source" tab. If your expecting an error there or on other destinations, you could try:
Adding a try-catch code block in your filter and transformer.
Use your custom code template function in your filter and transformer to catch the error or issue.
Create a separate channel that will receive the ACK/NACK which will be responsible in forwarding that message to the client.
In your try-catch code block or custom code template, use the method router.routeMessageByChannelId to forward the ACK/NACK to the other channel (step 3).
Note: You'll need to disable the default response in your original channel since you have the other channel that'll forward the ACK/NACK. You'll also need to consider if the client's machine expects a valid ACK/NACK immediately when they sent the HL7 message, depends on their setup.

Azure Service Bus Queue gives control characters in fetched message

I am using BizTalk Server SB-Messaging adapter to retreive messages from Azure Service Bus Queue. I have successfully managed to send message to queue myself (using same adapter), and retreive message from queue and do further processing.
Problem arises when a 3rd party software supplier is sending messages to the queue, and for BizTalk Server to retreive and process message. I then receive the following additional "header"-information and control characters in the beginning of the message:
In text: #ACKstringBShttp://schemas.microsoft.com/2003/10/Serialization/?$SOH
Seems like there is some sort of enveloped message, including headers to handle ACKnowledgement of the message to the queue.
SB-Messaging adapter gave following initial error message:
"The WCF service host at address 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."
And, another error message:
"No Disassemble stage components can recognize the data."
Did anyone hit this problem before, and, what can be the cause of the problem? Can character encoding be a possible cause of this problem?
Here comes the feedback!
Turned out 3rd party software supplier had a setting to send message as stream, instead of string. Turns out it is a .Net application using BrokeredMessage object. Using string makes message serialized, and meta-data is added to the message. Using stream, no such serialization takes place, and message is kept untouched.
So, problem was using string and automatic serialization when sending to Service Bus Queue.
I have legacy Microsoft.ServiceBus.Messaging clients sending BrokeredMessage Xml content as <string> and I want to receive using the latest Microsoft.Azure.ServiceBus library and Message type.
Using Encoding.UTF8.GetString(message.Body) I get a unusable string prefaced with
#\u0006string\b3http://schemas.microsoft.com/2003/10/Serialization/��#
My approach is to explicitly use XmlDictionaryReader binary deserialization to undo the hidden serialization magic from the legacy library
private static string GetMessageBodyHandlingLegacyBrokeredMessage(Message message)
{
string str = null;
if (message.ContentType == null)
{
using (var reader = XmlDictionaryReader.CreateBinaryReader(
new MemoryStream(message.Body),
null,
XmlDictionaryReaderQuotas.Max))
{
var doc = new XmlDocument();
doc.Load(reader);
str = doc.InnerText;
}
}
else
throw new NotImplementedException("Unhandled Service Bus ContentType " + message.ContentType);
return str;
}
References
https://learn.microsoft.com/en-us/azure/service-bus-messaging/service-bus-messages-payloads#payload-serialization
https://carlos.mendible.com/2016/07/17/step-by-step-net-core-azure-service-bus-and-amqp/

Intercept/Callback for QuickFIX message

I am using a FIX protocol to communicate with one of our counterparties. I have used Camel with Spring to build my communication routes.
I have a requirement where in my counterparty is expecting an ACK for every request it sends to me.
For example:
TradeCaptureRequestAck in response to TradeCaptureRequest
AllocationReportAck in response to AllocationReport
Confirmation_Ack in response to Confirmation
They are expecting a response irrespective of what happens at our end (even if something fails or exception occurs).
One way I know we can intercept the incoming message via MessageFactory. We can create a custom messagefactory and inject it in while creating QuickFixJComponent bean.
Problem with this approach is at factory level I will just be able to get the message type like TradeCaptureReport, AllocationReport etc. but not the content because factory only creates (and returns) the appropriate Message object. Actual work of populating this message object with incoming message data happens in Session class I guess (not sure about this).
Can someone please tell me if there is a way I can get or intercept the request message as soon as it reaches the route so that I can send the appropriate ACK to counterparty?