I'm trying to understand sagas, and meanwhile I have a specific way of thinking of them - but I am not sure whether I got the idea right. Hence I'd like to elaborate and have others tell me whether it's right or wrong.
In my understanding, sagas are a solution to the question of how to model long-running processes. Long-running means: Involving multiple commands, multiple events and possibly multiple aggregates. The process is not modeled inside one of the participating aggregates to avoid dependencies between them.
Basically, a saga is nothing more but a command / event handler that reacts on internal and external commands / events. It does not contain its own logic, it's just a (finite) state machine, and therefor provides tasks such as When event X happens, send command Y.
Sagas are persisted to the event store as well as aggregates, are correlated to a specific aggregate instance, and hence are reloaded when this specific aggregate (or set of aggregates) is used.
Is this right?
There are different means of implementing Sagas. Reaching from stateless event handlers that publish commands all the way to carrying all the state and basically being the domain's aggregates themselves. Udi Dahan once wrote an article about Sagas being the only Aggregates in a (in his specific case) correctly modeled system. I'll look it up and update this answer.
There's also the concept of document-based sagas.
Your definition of Sagas sounds right for me and I also would define them so.
The only change in your description I would made is that a saga is only a eventhandler (not a command) for event(s) and based on the receiving event and its internal state constructs a command and sents it to the CommandBus for execution.
Normally has a Saga only a single event to be started from (StartByEvent) and multiple events to transition (TransitionByEvent) to the next state and mutiple event to be ended by(EndByEvent).
On MSDN they defined Sagas as ProcessManager.
The term saga is commonly used in discussions of CQRS to refer to a
piece of code that coordinates and routes messages between bounded
contexts and aggregates. However, for the purposes of this guidance we
prefer to use the term process manager to refer to this type of code
artifact. There are two reasons for this: There is a well-known,
pre-existing definition of the term saga that has a different meaning
from the one generally understood in relation to CQRS. The term
process manager is a better description of the role performed by this
type of code artifact. Although the term saga is often used in the
context of the CQRS pattern, it has a pre-existing definition. We have
chosen to use the term process manager in this guidance to avoid
confusion with this pre-existing definition. The term saga, in
relation to distributed systems, was originally defined in the paper
"Sagas" by Hector Garcia-Molina and Kenneth Salem. This paper proposes
a mechanism that it calls a saga as an alternative to using a
distributed transaction for managing a long-running business process.
The paper recognizes that business processes are often comprised of
multiple steps, each of which involves a transaction, and that overall
consistency can be achieved by grouping these individual transactions
into a distributed transaction. However, in long-running business
processes, using distributed transactions can impact on the
performance and concurrency of the system because of the locks that
must be held for the duration of the distributed transaction.
reference: http://msdn.microsoft.com/en-us/library/jj591569.aspx
Related
Consider the simple use case in which I want to store product ratings as events in an event store.
I could use two different approaches:
Using Axon: A Rating aggregate is responsible for handling the CreateRatingCommand and sending the RatingCreatedEvent. Sending the event would case the Rating to be stored in the event store. Other event handlers have the possibility to replay the event stream when connecting to the Axon server instance and doing whatever needed with the ratings. In this case, the event handler will be used as a stream processor.
Using Kafka: A KafkaProducer will be used to store a Rating POJO (after proper serialization) in a Kafka topic. Setting the topic's retention time to indefinite would cause no events to get lost in time. Kafka Streams would in this case be used to do the actual rating processing logic.
Some architectural questions appear to me for both approaches:
When using Axon:
Is there any added value to use Axon (or similar solutions) if there is no real state to be maintained or altered within the aggregate? The aggregate just serves as a "dumb" placeholder for the data, but does not provide any state changing logic.
How does Axon handle multiple event handlers of the same event type? Will they all handle the same event (same aggregate id) in parallel, or is the same event only handled once by one of the handlers?
Are events stored in the Axon event store kept until the end of time?
When using Kafka:
Kafka stores events/messages with the same key in the same partition. How does one select the best value for a key in the use case of user-product ratings? UserId, ProductId or a separate topic for both and publish each event in both topics.
Would it be wise to use a separate topic for each user and each product resulting in a massive amount of topics on the cluster? (Approximately <5k products and >10k users).
I don't know if SO is the preferred forum for this kind of questions... I was just wondering what you (would) recommend in this particular use case as the best practise. Looking forward to your feedback and feel free to point out other points of thought I missed in the previous questions.
EDIT#12/11/2020 : I just found a related discussion containing useful information related to my question.
As Jan Galinski already puts it, this hasn't got a fool proof answer to it really. This is worth a broader discussion on for example indeed AxonIQ's Discuss forum. Regardless, there are some questions in here I can definitely give an answer to, so let's get to it:
Axon Question 1 - Axon Framework is as you've noticed used a lot for DDD centric applications. Nothing however forces you to base yourself on that notion at all. You can strip the framework from Event Sourcing specifics, as well as modelling specifics entirely and purely go for the messaging idea of distinct commands, events and queries. It has been a conscious decision to segregate Axon Framework version 3 into these sub-part when version 4 (current) was released actually. Next to that, I think there is great value in not just basing yourself on event messages. Using distinct commands and queries only further decouples your components, making for a far richer and easier to extend application landscape.
Axon Question 2 - This depends on where the #EventHandler annotated methods are located actually. If they're in the same class only one will be invoked. If they're positioned into distinct classes, then both will receive the same event. Furthermore if they're segregated between distinct classes, it is important to note Axon uses an Event Processor as the technical solution to invoking your event handlers. If distinct classes are grouped under the same Event Processor, you can impose a certain ordering which handler is invoked first. Next to this if the event handling should occur in parallel, you will have to configure a so called TrackingEventProcessor (the default in Axon Framework), as it allows configuration of several threads to handle events concurrently. Well, to conclude this section, everything you're asking in question two is an option, neither a necessity. Just a matter of configuration really. Might be worth checking up on this documentation page of Axon Framework on the matter.
Axon Question 3 - As Axon Server serves the purpose of an Event Store, there is no retention period at all. So yes, they're by default kept until the end of time. There is nothing stopping your from dropping the events though, if you feel there's no value in storing the events to for example base all your models on (as you'd do when using Event Sourcing).
It's the Kafka question I'm personally less familiar with (figures as a contributor to Axon Framework I guess). I can give you my two cents on the matter here too though, although I'd recommend a second opinion here:
Kafka Question 1 - From my personal feeling of what such an application would require, I'd assume you'd want to be able to retrieve all data for a given product as efficient as possible. I'd wager it's important that all events are in the same partition to make this process as efficient as possible, is it wouldn't require any merging afterwards. With this in mind, I'd think using the ProductId will make most sense.
Kafka Question 2 - If you are anticipating only 5_000 products and 10_000 users, I'd guess it should be doable to have separate topics for these. Opinion incoming - It is here though were I personally feel that Kafka's intent to provide you direct power to decide on when to use topics over complicates from what you'd actually try to achieve, which business functionality. Giving the power to segregate streams feels more like an after thought from the perspective of application development. As soon as you'd require an enterprise grade/efficient message bus, that's when this option really shines I think, as then you can optimize for bulk.
Hoping all this helps you further #KDW!
Experts,
I am in the evaluating moving a nice designed DDD app to an event sourced architecture as a pet project.
As can be inferred, my Aggregate operations are relatively coarse grained. As part of my design process, I now find myself emitting a large number of events from a small subset of operations to satisfy what I know to be requirements of the read model. Is this acceptable practice ?
In addition to this, I have distilled a lot of the domain complexity via use of ValueObjects & entities. Can VO's/ E accept commands and emit events themselves, or should I expose state and add from the command handler lower down the stack ?
In terms of VO's note that I use mutable operations sparingly and it is a trade off between over complicating other areas of my domain.
I now find myself emitting a large number of events from a small subset of operations to satisfy what I know to be requirements of the read model. Is this acceptable practice ?
In most cases, you should have 1 event by command. Remember events describe the user intent so you have to stay close to the user action.
Can VO's/ E accept commands and emit events themselves
No only aggregates emit events and yes you can come up with very messy aggregates if you have a lot of events, but there is solutions for that like delegating the work to the commands and events themselves. I blogged about this technique here: https://dev.to/maximegel/cqrs-scalable-aggregates-731
In terms of VO's note that I use mutable operations sparingly
As long as you're aware of the consequences. That's fine. Trade-off are part of the job, but be sure you team is aware of that since it's written everywhere that value objects are immutable you expose yourselves to confusion and pointer issues.
Assuming a BPMN process describing activities, gateways, start and end events. As follow:
Each step is managed by a BPMN engine. At one point, how can we tell which is the state of the process ? Activities seem to define some state embodied as actions (e.g. evaluating request). Am I correct ?
Also, if we assume activity represents the state, how do we get a listing of next possible states if we were to navigate through a dedicated follow-up application ?
Should the process be modeled in a more workflow oriented way to express those state/actions possibilities ? I have the intuition that events could also be used to manage states and possible related actions.
Since I am not sure what exactly you understand as state of the process, I will try to define that first. I guess you are aware of the token concept, see a discussion in the Camunda forum:
A token is a BPMN concept that represents a state within a process instance. It does not have any variables or any message.
You may now define the state of the process as a statistics how many tokens at a given time are existing, and how many are currently in a given activity or event.
This statistics can be extracted from your favorite BPMN engine (and seen e.g. in Camunda's Cockpit as little colorful bubbles). With that statistics in hand, you could in principle generate forecast on next possible states, i.e. determine scenarios how many tokens will be in the next time instance probably in each activity.
State has a different meaning in BPMN, it could mean:
1 - Where is the token in the flow?
2 - Is the process flow running correctly or not?
3 - Or, by a specific variable (field) in the forms.
If you mean the third case, which is common in processes, you have to define a field in your data model as enum (depends on the engine) and manually or automatically change its value in the forms.
Obviously, the rather abstract Petri-Net-style token flow semantics of BPMN does not capture the real semantics of business processes. It has just been artificially imposed on BPMN due to academic pressure groups. A really meaningful semantics must refer to the information context of a process in the business system that owns it.
Of course, a business system that is the owner of a process (type), is, at any point during a running process, in a certain complex dynamic information state, some part of which forms the context of the process and can therefore be considered its state.
In fact, the (information) state of a process is essentially given by all the property-value slots of objects that are used or affected by (events/activities of) the process. In addition to these "global variables", the state of a process also includes
the values of (auxiliary) process variables,
the information, which activities have been started (and are ongoing).
Take a look into the Imixs-Workflow project. It is a event orientated workflow engine instead of the task orientated design often seen in BPM engines.
Each task in this kind of workflow engine defines a state in your process model. The workflow engine holds this state until a event is fired. An event defines the transition from one state to another.
You can find examples how to model different szenarious in a event driven workflow model here.
I'm wandering on DDD and NoSql field actually. I have a doubt now: i need to produce events from the aggregate and i would like to use a NoSql storage. But how can i be sure that events are saved on the storage AND the changes on the aggregate root not having transactions?
Does it makes sense? Is there a way to do this without being forced to use event sourcing or a transactional db?
Actually i was lookin at implementing a 2 phase commit algorithm but it seems pretty heavy from a performance point of view...
Am i approaching the problem the wrong way?
Stuffed with questions...
Thanks for every suggestion
Enrico
PS
I'm a newbie on stackoverflow so any suggestion/critic/... is more than welcome
Enrico
Edit 1
Well i would need events to notify aggregates that something happened and i they should react to the change. The problem arise when such events are important for the business logic. As far as i understood, after a night of thinking, i can't use a nosql storage to do such things. Let me explain (thinking with loud voice :P):
With ES (1st scenery): I save the "diff" of the data. Then i produce an event associated with it. 2 operations.
With ES (2nd scenery): I save the "diff" of the data. A process, watch the ES and produce the event. But i'm tied to having only one watcher process to ensure the correct ordering of events.
With ES (3d scenery): Idempotent events. The events can be inferred by the state and every reapplication of the event can cause a change on the consumer only once, can have multiple "dequeue" processes, duplicates can't possibly happen. 1 operation, but it introduce heavy limitations on the consumers.
In general: I save the aggregate's data. Then i produce an event associated with it. 2 operations.
Now the question becomes wider imho, is it possible to work with domain events and nosql when such domain events are fundamental part of the business process?
I think that could be a better option to go relational... even if i would need to add quite a lot of machines to get the same performances.
Edit 2
For the sake of completness, searching for "domain events nosql idempotent" on google: http://svendvanderveken.wordpress.com/2011/08/26/transactional-event-based-nosql-storage/
If you need Event Sourcing, you should store events only.
This should be the sequence:
the aggregate root recieves a command
it fires proper events
events are stored
Each aggregate's re-hydratation should be done only by executing events over them. You can create aggregates' snapshots if you measure performance problems on their initialization, but this doesn't require two-phase commits, since you can build snapshots asynchronously via batch.
Note however that you need CQRS and/or Event Sourcing only if your application is heavily concurrent and you need to cope with partition tolerance and compensating actions.
edit
Event Sourcing is alternative to the persistence of object state. You either store the events or the state of the object model. You can save snapshot, but they're just performance tools: your application must be able to work without them. You can consider such snapshots as a caching technique. As an alternative you can persist object state (the classical model), but in that case you don't need to store events.
In my own DDD application, I use observable entities to decouple (via direct events' subscription from the repository) aggregates and their persistence. For example your repository can subscribe each domain events, and execute the actions required by the application (persist to the store, dispatch to a queue and so on...). But as a persistence technique, Event Sourcing is alternative to classical persistence of the observable object state. In most scenarios you don't need both.
edit 2
A final note: if you choose ES, one of the events subscriber can build a relational read-model too.
Didn't know how to shorten that title.
I'm basically trying to wrap my head around the concept of CQRS (http://en.wikipedia.org/wiki/Command-query_separation) and related concepts.
Although CQRS doesn't necessarily incorporate Messaging and Event Sourcing it seems to be a good combination (as can be seen with a lot of examples / blogposts combining these concepts )
Given a use-case for a state change for something (say to update a Question on SO), would you consider the following flow to be correct (as in best practice) ?
The system issues an aggregate UpdateQuestionCommand which might be separated into a couple of smaller commands: UpdateQuestion which is targeted at the Question Aggregate Root, and UpdateUserAction(to count points, etc) targeted at the User Aggregate Root. These are send asynchronously using point-to-point messaging.
The aggregate roots do their thing and if all goes well fire events QuestionUpdated and UserActionUpdated respectively, which contain state that is outsourced to an Event Store.. to be persisted yadayada, just to be complete, not really the point here.
These events are also put on a pub/sub queue for broadcasting. Any subscriber (among which likely one or multiple Projectors which create the Read Views) are free to subscribe to these events.
The general question: Is it indeed best practice, that Commands are communicated Point-to-Point (i.e: The receiver is known) whereas events are broadcasted (I.e: the receiver(s) are unknown) ?
Assuming the above, what would be the advantage/ disadvantage of allowing Commands to be broadcasted through pub/sub instead of point-to-point?
For example: When broadcasting Commands while using Saga's (http://blog.jonathanoliver.com/2010/09/cqrs-sagas-with-event-sourcing-part-i-of-ii/) could be a problem, since the mediation role a Saga needs to play in case of failure of one of the aggregate roots is hindered, because the saga doesn't know which aggregate roots participate to begin with.
On the other hand, I see advantages (flexibility) when broadcasting commands would be allowed.
Any help in clearing my head is highly appreciated.
Yes, for Command or Query there is only one and exactly one receiver (thus you can still load balance), but for Events there could be zero or more receivers (subscribers)