Problem
Suppose there is a Http request observable that errored, we can just retry it. But I also want the UI to inform the user that this resource failed to load. What is the best architecture?
Intended Behavior for the Target Observable
Retry-able.
Long-running. Doesn't complete or error.
Shared. Does not generate unnecessary requests when multiple subscriber.
Load on need. Does not generate unnecessary requests when not subscribed.
Inform UI of the errors.
(3 and 4 can be achieved by shareReplay({bufferSize: 1, refCount: true}))
My Attempts
I think it's best to pass an error message to the downstream observer while keeping retrying the source. It causes minimum changes to the architecture. But I didn't see a way I can do it with Rxjs, because
retry() always intercepts the error. If you materialze the error, then retry() won't retry. If not, then no error will propagate to the downstream.
catchError() without rethrowing will always complete the stream.
Although let the UI observer tap(,,onError) and retry() can satisfy this need, but I think it is dangerous to let the UI take this responsibility. And multiple UI observer means a LOT of duplicated retries.
Well, I seem to have accidentally find the answer while browsing through the documentations.
It starts with the usage of the second parameter of the catchError. According to the documentation, retry is implemented by catchError. And we can express more logic with the lower-level catchError.
So it's just
catchError((err, caught) => {
return timer(RETRY_DELAY_TIME).pipe(
mergeMap(() => caught)
startWith(err)
);
})
It retries the observable, meanwhile sending error messages to the downstream observers. So the downstream is aware of the connection error, and can expect to receive retried values.
It sounds like you're looking for something akin to an NgRx side effect. You can encase it all in an outer Observable, piping the error handler to the inner Observable (your HTTP call), something like this:
const myObs$ = fromEvent('place event that triggers call here').pipe(
// just one example, you can trigger this as you please
switchMap(() => this.myHttpService.getResource().pipe(
catchError(err => handleAndRethrowError()),
retry(3)
),
shareReplay()
);
This way, if the request throws an error, it is retried 3 times (with error handling in the catchError block, and even if it fully errors out, the outer Observable is still alive. Does that look like it makes sense?
Related
Project Reactor has a variety of timeout() operators.
The very basic implementation raises TimeoutException in case no item arrives within the given Duration. The exception is propagated downstream , and to upstream it sends cancel signal.
Basically my question is: is it possible to somehow react (and do something) specifically to timeout that happened downstream, not just to cancelation that sent after timeout happened?
My question is based on the requirements of my real business case and also I'm wondering if there is a straight solution.
I'll simplify my code for better understanding what I want to achieve.
Let's say I have the following reactive pipeline:
Flux.fromIterable(List.of(firstClient, secondClient))
.concatMap(Client::callApi) // making API calls sequentially
.collectList() // collecting results of API calls for further processing
.timeout(Duration.ofMillis(3000)) // the entire process should not take more than duration specified
.subscribe();
I have multiple clients for making API calls. The business requirement is to call them sequantilly, so I call them with concatMap(). Then I should collect all the results and the entire process should not take more than some Duration
The Client interface:
interface Client {
Mono<Result> callApi();
}
And the implementations:
Client firstClient = () ->
Mono.delay(Duration.ofMillis(2000L)) // simulating delay of first api call
.map(__ -> new Result())
// !!! Pseudo-operator just to demonstrate what I want to achieve
.doOnTimeoutDownstream(() ->
log.info("First API call canceled due to downstream timeout!")
);
Client secondClient = () ->
Mono.delay(Duration.ofMillis(1500L)) // simulating delay of second api call
.map(__ -> new Result())
// !!! Pseudo-operator just to demonstrate what I want to achieve
.doOnTimeoutDownstream(() ->
log.info("Second API call canceled due to downstream timeout!")
);
So, if I have not received and collected all the results during the amount of time specified, I need to know which API call was actually canceled due to downstream timeout and have some callback for this "event".
I know I could put doOnCancel() callback to every client call (instead of pseudo-operator I demonstrated) and it would work, but this callback reacts to cancelation, which may happen due to any error.
Of course, with proper exception handling (onErrorResume(), for example) it would work as I expect, however, I'm interesting if there is some straight way to somehow react specifically to timeout in this case.
I want to make sure that all events, which were submitted before complete was invoked on the observable, are logged.
I'm aware that operators exist that stop emission of events (takeUntil, etc.) completely.
The question that I have is whether other operators exist which would lead to emissions not being sent if the complete on the subject is sent too 'early'. Are there cases where it would be beneficial to wait with the completion of the observable until the event was handled by the observer?
For example, are there situations (imagine any other RxJS operator instead of the delay) where the following code ...
const subj = new Subject<string>();
subj.pipe(delay(500))
.subscribe((val) => {
console.log(val);
subj.complete();
});
subj.next('1');
... makes more sense than that ...
const subj = new Subject<string>();
subj.pipe(delay(500))
.subscribe((val) => {
console.log(val);
});
subj.next('1');
subj.complete();
... when the subject should only emit one value?
Or is completing the subject immediately after next always safe in such situations?
If there are other factors I'm not aware of (e.g. synchronous vs. asynchronous execution of code) please mention them as well.
In general,
subj.next('1');
subj.complete();
is safe. As far as I know, none of the existing RxJS operators use a completion to cancel/unsubscribe observables early.
That being said, I can easily create such an operator myself. switchMap will cancel currently running inner observables when it receives it's next emission. You could, for example, create a custom operator that unsubscribes and exists early when it receives a complete event.
If your worried about that, however, you're out of luck. No matter what workaround you imagine, I can probably write an operator that will ruin your code. For example filter(_ => false) will stop the 1 from being emitted before the complete in either of the two cases you've described.
In the end, you and your coworkers must write good code (and test!) and RxJS doesn't change that.
I am using the following code to basically ensure that I get a result from an Observable
this.authenticationService.isLoggedIn
.pipe(
take(1),
defaultIfEmpty(false)
)
.subscribe(result => return result);
Somehow it feels wrong to me, maybe because it seems sort of procedural.
Is this method okay? Will this get me in trouble in any way?
If in your code it's fine that this.authenticationService.isLoggedIn completes without an emit - then the code in your question is fine too.
If this.authenticationService.isLoggedIn emits anything at some point of time and completes after - then defaultIfEmpty is redundant.
It all depends on what isLoggedIn does.
It is clear that isLoggedIn returns an Observable.
Now, and Observable can do just 3 things
it can notify, i.e. emit, some data for consumption of Observers which are subscribed
it can raise an error
it can complete
So the first question is: how many times can isLoggedIn notify? Is it just one shot or is it a stream of notifications? If it can emit just one value and then complete, than the take(1) operator is useless.
But there is also the case that isLoggedIn never notifies and just completes. In this case a notification would never be signaled by the observer to its subscriber. Using defaultIfEmpty operator ensures that something is notified even in this case.
So, reading your code I understand that isLoggedIn can behave in these 2 ways
Emit more than once but you are interested only in the first notification
Never notify and just complete, in which case you want false to be returned
If this is not true, it may be the case that your code can be simplified.
This is a bit of a conceptual question, so let me know if it's off topic.
I'm looking at writing yet another library to process messages off a queue - in this case an Azure storage queue. It's pretty easy to create an observable and throw a message into it every time a message is available.
However, there's a snag here that I'm not sure how to handle. The issue is this: when you're done processing the message, you need to call an API on the storage queue to actually delete the message. Otherwise the visibility timeout will expire and the message will reappear to be dequeued again.
As an example, here's how this loop looks in C#:
public event EventHandler<string> OnMessage;
public void Run()
{
while(true)
{
// Read message
var message = queue.GetMessage();
if (message != null)
{
// Run any handlers
OnMessage?.Invoke(this, message.AsString);
// Delete off queue when done
queue.DeleteMessage(message);
}
else
{
Thread.Sleep(2500);
}
}
}
The important thing here is that we read the message, trigger any registered event handlers to do things, then delete the message after the handlers are done. I've omitted error handling here, but in general if the handler fails we should NOT delete the message, but instead let it return to visibility automatically and get redelivered later.
How do you handle this kind of thing using Rx? Ideally I'd like to expose the observable for anyone to subscribe to. But I need to do stuff at the end of processing for that message, whatever the "end" happens to mean here.
I can think of a couple of possible solutions, but I don't really like any of them. One would be to have the library call a function supplied by the consumer, that takes in the source observable, hooks up whatever it wants, then returns a new observable that the library can then subscribe on to do the final cleanup. But that's pretty limiting, as consumers basically only have one shot to hook up to the messages, which seems pretty limiting.
I guess I could put the call to delete the message after the call to onNext, but then I don't know if the processing succeeded or failed unless there's some sort of back channel in that api I don't know about?
Any ideas/suggestions/previous experience here?
Try having a play with this:
IObservable<int> source =
Observable
.Range(0, 3)
.Select(x =>
Observable
.Using(
() => Disposable.Create(() => Console.WriteLine($"Removing {x}")),
d => Observable.Return(x)))
.Merge();
source
.Subscribe(x => Console.WriteLine($"Processing {x}"));
It produces:
Processing 0
Removing 0
Processing 1
Removing 1
Processing 2
Removing 2
In short:
How to proceed listening after an error in stream without putting a .catch before every .subscribe?
If you need more details they are here:
Lets assume I have a Subject of current user or null. I get the data from API sometimes and send to the Subject. It updates the view accordingly.
But at some point error occurs on my server and I want my application to continue working as before but notify some places about the error and KEEP listening to my Subject.
Initially I thought that if I just do userSubject.error(...) it will only trigger .catch callback and error handlers on subscribes and skip all success handlers and chains.
And if after I call userSubject.next(...) all my chains and subscribers will work as before
BUT unluckily it is not the case. After the first uncaught .error it unsubscribes subscribers from the stream and they do not operate any more.
So my question: Why???
And what to do instead if I want to handle null value normally but also handle errors only in some places?
Here is the link to RxJs source code where Subscriber unsubscribes on error
https://github.com/ReactiveX/rxjs/blob/master/src/Subscriber.ts#L140
Rx observables follow the grammar next*(error|complete)?, meaning that they can produce nothing after error or complete notification has been delivered.
An explanation of why this matters can be found from Rx design guidelines:
The single message indicating that an observable sequence has finished ensures that consumers of the observable sequence can deterministically establish that it is safe to perform cleanup operations.
A single failure further ensures that abort semantics can be maintained for operators that work on multiple observable sequences.
In short, if you want your observers to keep listening to the subject after a server error has occurred, do not deliver that error to the subject, but rather handle it in some other way (e.g. use catch, retry or deliver the error to a dedicated subject).
Every Observable emits zero or more next notifications and one error or complete but never both.
For this reason, Subjects have internal state.
Then it depends how you construct your chain. For example you can use retry() to resubscribe to its source Observable on error.
Or when you pass values to your Subject you can send only next notifications and ignore the other two:
.subscribe(v => subject.next(v));
Or if you want to throw error when the user is null you can use any operator that captures exceptions and sends them as error notifications. For example like this:
.map(v => {
if (v === null) {
throw new Error("It's broken");
}
return v;
})
Anyway it's hard to give more precise advice without any code.