In the last article Workflow analysis of Hystrix In, we introduced the workflow of Hystrix as a whole, knowing that Hystrix will degrade in the following four situations:

1. Fuse open
2. Thread pool / semaphore full
3. Call timeout
4. Call failed

This article introduces the handling details of Hystrix in the event of degradation. The following method exception handling logic:

final Func1<Throwable, Observable<R>> handleFallback = new Func1<Throwable, Observable<R>>() {
            @Override
            public Observable<R> call(Throwable t) {
                circuitBreaker.markNonSuccess();
                Exception e = getExceptionFromThrowable(t);
                executionResult = executionResult.setExecutionException(e);
                if (e instanceof RejectedExecutionException) {
                    return handleThreadPoolRejectionViaFallback(e);
                } else if (t instanceof HystrixTimeoutException) {
                    return handleTimeoutViaFallback();
                } else if (t instanceof HystrixBadRequestException) {
                    return handleBadRequestByEmittingError(e);
                } else {
                    /*
                     * Treat HystrixBadRequestException from ExecutionHook like a plain HystrixBadRequestException.
                     */
                    if (e instanceof HystrixBadRequestException) {
                        eventNotifier.markEvent(HystrixEventType.BAD_REQUEST, commandKey);
                        return Observable.error(e);
                    }

                    return handleFailureViaFallback(e);
                }
            }
        };

Here we take the logic that the thread pool is full to analyze

Thread pool full

    private Observable<R> handleThreadPoolRejectionViaFallback(Exception underlying) {
        eventNotifier.markEvent(HystrixEventType.THREAD_POOL_REJECTED, commandKey);
        threadPool.markThreadRejection();
        // use a fallback instead (or throw exception if not implemented)
        return getFallbackOrThrowException(this, HystrixEventType.THREAD_POOL_REJECTED, FailureType.REJECTED_THREAD_EXECUTION, "could not be queued for execution", underlying);
    }

1. The first line publishes an event rejected by the thread pool
2. The second line records the number of thread pool rejections
3. Get Fallback method

Get Fallback method

    final HystrixRequestContext requestContext = HystrixRequestContext.getContextForCurrentThread();
        long latency = System.currentTimeMillis() - executionResult.getStartTimestamp();
        // record the executionResult
        // do this before executing fallback so it can be queried from within getFallback (see See https://github.com/Netflix/Hystrix/pull/144)
        executionResult = executionResult.addEvent((int) latency, eventType);

        if (isUnrecoverable(originalException)) {
            logger.error("Unrecoverable Error for HystrixCommand so will throw HystrixRuntimeException and not apply fallback. ", originalException);

            /* executionHook for all errors */
            Exception e = wrapWithOnErrorHook(failureType, originalException);
            return Observable.error(new HystrixRuntimeException(failureType, this.getClass(), getLogMessagePrefix() + " " + message + " and encountered unrecoverable error.", e, null));
        } else {
            if (isRecoverableError(originalException)) {
                logger.warn("Recovered from java.lang.Error by serving Hystrix fallback", originalException);
            }

            if (properties.fallbackEnabled().get()) {
                /* fallback behavior is permitted so attempt */

                final Action1<Notification<? super R>> setRequestContext = new Action1<Notification<? super R>>() {
                    @Override
                    public void call(Notification<? super R> rNotification) {
                        setRequestContextIfNeeded(requestContext);
                    }
                };

                final Action1<R> markFallbackEmit = new Action1<R>() {
                    @Override
                    public void call(R r) {
                        if (shouldOutputOnNextEvents()) {
                            executionResult = executionResult.addEvent(HystrixEventType.FALLBACK_EMIT);
                            eventNotifier.markEvent(HystrixEventType.FALLBACK_EMIT, commandKey);
                        }
                    }
                };

                final Action0 markFallbackCompleted = new Action0() {
                    @Override
                    public void call() {
                        long latency = System.currentTimeMillis() - executionResult.getStartTimestamp();
                        eventNotifier.markEvent(HystrixEventType.FALLBACK_SUCCESS, commandKey);
                        executionResult = executionResult.addEvent((int) latency, HystrixEventType.FALLBACK_SUCCESS);
                    }
                };

                final Func1<Throwable, Observable<R>> handleFallbackError = new Func1<Throwable, Observable<R>>() {
                    @Override
                    public Observable<R> call(Throwable t) {
                        /* executionHook for all errors */
                        Exception e = wrapWithOnErrorHook(failureType, originalException);
                        Exception fe = getExceptionFromThrowable(t);

                        long latency = System.currentTimeMillis() - executionResult.getStartTimestamp();
                        Exception toEmit;

                        if (fe instanceof UnsupportedOperationException) {
                            logger.debug("No fallback for HystrixCommand. ", fe); // debug only since we're throwing the exception and someone higher will do something with it
                            eventNotifier.markEvent(HystrixEventType.FALLBACK_MISSING, commandKey);
                            executionResult = executionResult.addEvent((int) latency, HystrixEventType.FALLBACK_MISSING);

                            toEmit = new HystrixRuntimeException(failureType, _cmd.getClass(), getLogMessagePrefix() + " " + message + " and no fallback available.", e, fe);
                        } else {
                            logger.debug("HystrixCommand execution " + failureType.name() + " and fallback failed.", fe);
                            eventNotifier.markEvent(HystrixEventType.FALLBACK_FAILURE, commandKey);
                            executionResult = executionResult.addEvent((int) latency, HystrixEventType.FALLBACK_FAILURE);

                            toEmit = new HystrixRuntimeException(failureType, _cmd.getClass(), getLogMessagePrefix() + " " + message + " and fallback failed.", e, fe);
                        }

                        // NOTE: we're suppressing fallback exception here
                        if (shouldNotBeWrapped(originalException)) {
                            return Observable.error(e);
                        }

                        return Observable.error(toEmit);
                    }
                };

                final TryableSemaphore fallbackSemaphore = getFallbackSemaphore();
                final AtomicBoolean semaphoreHasBeenReleased = new AtomicBoolean(false);
                final Action0 singleSemaphoreRelease = new Action0() {
                    @Override
                    public void call() {
                        if (semaphoreHasBeenReleased.compareAndSet(false, true)) {
                            fallbackSemaphore.release();
                        }
                    }
                };

                Observable<R> fallbackExecutionChain;

                // acquire a permit
                if (fallbackSemaphore.tryAcquire()) {
                    try {
                        if (isFallbackUserDefined()) {
                            executionHook.onFallbackStart(this);
                            fallbackExecutionChain = getFallbackObservable();
                        } else {
                            //same logic as above without the hook invocation
                            fallbackExecutionChain = getFallbackObservable();
                        }
                    } catch (Throwable ex) {
                        //If hook or user-fallback throws, then use that as the result of the fallback lookup
                        fallbackExecutionChain = Observable.error(ex);
                    }

                    return fallbackExecutionChain
                            .doOnEach(setRequestContext)
                            .lift(new FallbackHookApplication(_cmd))
                            .lift(new DeprecatedOnFallbackHookApplication(_cmd))
                            .doOnNext(markFallbackEmit)
                            .doOnCompleted(markFallbackCompleted)
                            .onErrorResumeNext(handleFallbackError)
                            .doOnTerminate(singleSemaphoreRelease)
                            .doOnUnsubscribe(singleSemaphoreRelease);
                } else {
                   return handleFallbackRejectionByEmittingError();
                }
            } else {
                return handleFallbackDisabledByEmittingError(originalException, failureType, message);
            }
        }
    }

The method is relatively long, mainly doing the following things:

1. Directly look at the isUnrecoverable method to determine whether the exception is an unrecoverable exception. If it is unrecoverable, it will directly return failure
2. Print log if recoverable exception
3. Judge whether to enable the execution fallback method. If it is enabled, go to step 4
4. Two events need to be sent to create start and finish: fallback? EIT, fallback? Success
5. When an exception occurs in the call fallback method, the handling logic is handleFallbackError. There are only two exceptions in this scenario:
   1. Unsupported operationexception: getFallback abstract method not implemented
   2. Other anomalies
6. Create Action to release semaphore: single semaphorerelease
7. Obtain the semaphore. If the fallback logic is executed successfully, that is, call the getFallback method implemented by the user

final protected Observable<R> getFallbackObservable() {
        return Observable.defer(new Func0<Observable<R>>() {
            @Override
            public Observable<R> call() {
                try {
                    return Observable.just(getFallback());
                } catch (Throwable ex) {
                    return Observable.error(ex);
                }
            }
        });
    }

Students interested in handling other types of exceptions can continue to study based on Func1 handleFallback