I have a REST endpoint called CancelOrder. It comprises four steps (in order):
Cancel fulfilment (Calls a downstream service).
Cancel the quote (Calls a downstream service).
Update the state of the order to cancelled (In our local database).
This is a PUT operation and hence, I am trying to make it idempotent and fail-safe.
Scenario 1:
Just a single call:
Cancels the fulfilment, cancels the quote, updates the state. All's good.
A call is midway when a different call is received. Assume no pessimistic locking is present:
The state of the order has not been changed to 'cancelled' by the previous call yet, but the fulfilment has been cancelled. Now, when the second call tried to cancel the fulfilment, it returns an error.
The ideal way to handle the above scenario is making the API transactional by acquiring a write lock on the document on each call. But I don't want to do that.
How should I handle this scenario?
There are 2 ways (among a lot of other solutions) to deal with this scenario:
Solution A:
Add a new state in order as isCanceling. After server receives the first cancel request on an order, set this state as true. Once the cancel operation is finished, set this state as false.
If server receives another cancel request on the same order, but find its status is isCanceling, server would return 102 Processing to client, indicating the operation is in-progress.
Solution B:
same as step 1 in Solution A.
Everytime when server receives the cancel request (including the first one), a listener is added in that order's queue, waiting to be notified by event "Cancel-OK" or "Cancel-Fail".
If server receives cancel request on an order, but find its status is isCanceling, server would do nothing but just add corresponding listener in the above queue.
Once cancel operation is finished (success or fail), an event is fired. All the listeners in the queue will get the message, and HTTP response would be returned for all the previous pending HTTP requests.
Personally, I prefer Solution B.
Related
I want to test a worker verticle that receives requests over EventBus and sends the results also over EventBus. A single request may result in 0,1,2,... responses - in general cases we don't know how many responses we'll get.
The business logic is that requests are acked once the processing is complete, however the responses are sent in "fire and forget" manner - therefore we only know the responses were sent, not necessarily that they were delivered already.
I am writing a test for this verticle.
The test code is planned to be like this:
1. set up consumer for responses
2. send a request
3. wait until request is acked by the worker verticle
4. wait until consumer finishes validating the responses
The problem here is step 4 - in general case we don't know if there are still some responses in flight or not.
A brute force solution is obviously to wait some reasonable time - a few milliseconds is usually enough. However. I'd prefer something more conceptual.
A solution that comes to my mind is this:
send some request for which we know for sure that there would be a single response;
wait until the consumer receives the corresponding response.
That should work, but I dislike the fact that I pump two messages through the SUT instead of just a single one.
A different solution would be to send one extra response from test code, once we have a confirmation that the request was processed - but would it be considered to be the same sender? The EventBus only guarantees delivery order from the same sender, not from different ones. The test doesn't run in cluster mode, all operations are performed on the same machine, though not necessarily in the same thread.
Yet another solution would be to somehow check that EventBus is now empty, but as I understand, this is not possible.
Is there any other (better) solution?
The solution I would choose now (after half a year more experience with vertx/EventBus) is to send two messages.
The second message would get acked only after the processing of the first one is complete.
This would only work if you have a single consumer so that your two messages can't be processed in parallel.
The documentation of Subscription#cancel says that
Data may still be sent to meet previously signalled demand after calling cancel.
In which scenario would people expect the publisher to continue to send till previous signalled demand is met?
Also, if I don't want any new items to be sent after cancellation, what should I do?
Unless you are creating low level operators or Publishers, you don't have to worry about this.
In which scenario would people expect the publisher to continue to send till previous signalled demand is met?
None of the mainstream Reactive Streams libraries do that as they stop sending items eventually. RxJava 2 and Reactor 3 are pretty eager on this so you'd most likely have an extra item on a low-lever asynchronously issued cancellation. Akka Stream may signal more than that (last time I checked, they mix control and item signals and there is a configuration setting for max synchronous items per stream that can lead to multiple items being emitted before the cancellation takes effect).
Also, if I don't want any new items to be sent after cancellation, what should I do?
Depends on what you implement: a Publisher or a Subscriber.
In a Publisher the most eager method is to set a volatile boolean cancelled field and check that every time you are in some kind of emission loop.
In a Subscriber, you can have a boolean done field that is checked in each onXXX so that when you call Subscription.cancel() from onNext, any subsequent call will be ignored.
At the moment I have a single AWS EC2 instance which handles all incoming http client requests. It analyses each request and then decides which back end worker server should handle the request and then makes a http call to the chosen server. The back end server then responds when it has processed the request. The front end server will then respond to the client. The front end server is effectively a load balancer.
I now want to go to a Pub-Sub architecture instead of the front end server pushing the requests to the back end instances. The front end server will do some basic processing and then simply put the request into an SNS queue and the logic of which back end server should handle the request is left to the back end servers themselves.
My question is with this model what is the best way to have the back end servers notify the front end server that they have processed the request? Previously they just replied to the http request the front end server sent but now there is no direct request, just an item of work being published to a queue and a back end instance picking it off the queue.
Pubsub architectures are not well suited to responses/acknowledgements. Their fire-and-forget broadcasting pattern decouples publishers and the subscribers: a publisher does not know if or how many subscribers there are, and the subscribers do no know which publisher generated a message. Also, it can be difficult to guarantee sequence of responses, they won't necessarily match the sequence of messages due to the nature of network comms and handling of messages can take different amounts of time etc. So each message that needs to be acknowledge needs a unique ID that the subscriber can include in its response so the publisher can match a response with the message sent. For example:
publisher sends message "new event" and provides a UUID for the
event
many subscribers get the message; some may be the handlers for
the request, but others might be observers, loggers, analytics, etc
if only one subscriber handles the message (e.g. the first
subscriber to get a key from somewhere), that subscriber generates a
message "new event handled" and provides a UUID
the original
publisher, as well as any number of other subscribers, may get that
message;
the original publisher sees the ID is
in its cache as an unconfirmed message, and now marks it as
confirmed
if a certain amount of time passes without receiving a
confirmation with given ID, the original publisher republishes the
original message, with a new ID, and removes the old ID from cache.
In step 3, if many subscribers handled the message instead of just one, then it
less obvious how the original publisher should handle "responses": how does it
know how many subscribers handle the message, some could be down or
too busy to respond, or some may be in the process of responding by the time
the original publisher determines that "not enough handlers have
responded".
Publish-subscribe architectures should be designed to not request any response, but instead to check for some condition that should have happened as a result of the command being handled, such as a thumbnail having gotten generated (it can assume as a result of a handler of the message).
I have implemented a chain of executions and each execution will send a HTTP request to the server and does check if the response status is 2XX. I need to implement a synchronous model in which the next execution in the chain should only get triggered when the previous execution is successful i.e response status is 2xx.
Below is the snapshot of the execution chain.
feed(postcodeFeeder).
exec(Seq(LocateStock.locateStockExecution, ReserveStock.reserveStockExecution, CancelOrder.cancelStockExecution,
ReserveStock.reserveStockExecution, ConfirmOrder.confirmStockExecution, CancelOrder.cancelStockExecution)
Since gatling has asynchronous IO model, what am currently observing is the HTTP requests are sent to the server in an asynchronous manner by a number of users and there is no real dependency between the executions with respect to a single user.
Also I wanted to know for an actor/user if an execution in a chain fails due the check, does it not proceed with the next execution in the chain?
there is no real dependency between the executions with respect to a single user
No, you are wrong. Except when using "resources", requests are sequential for a given user. If you want to stop the flow for a given user when it encounters an error, you can use exitblockonfail.
Gatling does not consider the failure response from the previous request before firing next in chain. You may need to cover the entire block with exitBlockOnFail{} to block the gatling to fire next.
As you know, WSASend-completion-WSASend-...(repeat) guarantees sending data in order.
However, I heard that WSASend-WSASend-WSASend-completion-completion-completion ALSO GUARANTEES sending data in order, although the order of completion does not. Is that right?
Please reply. Thanks in advance.
Hyunjik Bae
If there are multiple requests queued up at a time, their data is sent in the order that they are queued, and their completion routines are queued up in the same order as well. The second send does not occur until the first send completes, the third send after the second send completes, etc. However, if you have multiple threads processing the completions, it is possible to have the completions being processed in parallel, which can get out of order depending on context switches and such. That's why you need to associate an identifying value with each pending request so you can match it with its completion regardless of the order, and don't have multiple requests pending for sending portions of the same resource.