I noticed that we can use Vert.x to write multiple Verticles and communicate using EventBus. Is this way different from writing some servers in just one Verticle?
You can create different servers in the same verticle, but all user requests will be handled by the same event loop.
This might work just fine for you use case. However usually it's best for clarity/performance to separate concerns.
Related
I have a microservices architecture whose logs have to be sent to a remote Kafka topic.
Next to it, the consumer of this topic will send the logs to an ELK stack (an other team)
I want to have a dedicated microservice (fwk-proxy-elasticsearch) whose responsability is to collec the logs from the others one and send them to the remote kafka topic.
what's the best protocol to dispatch all the logs aggregated from my microservices to the fwk-proxy-elasticsearch microservice ?
I want this pattern to not duplicate the security configuration of the remote kafka topic. I want to centralize it in a single place.
May I use vertx event bus for that ? or kafka is beter ? or someother tool ?
May I use vertx to send message from jvm to jvm ?
Moreover, in a microservice architecture, is it a good pattern to centralize a use case in a dedicated microservice? (remote http connection for example)
On my point of view, it allows business microservices to focus on a business issue and not to worry over the protocol that the result has to be sent.
Thanks!
I believe you can use both Vert.x event bus and Kafka to propagate the logs, there are pros and cons on each approach.
While I understand the reasoning behind this decision, I would still consider a dedicated solution built for this purpose, like Fluentd, which is able to aggregate the logs and push them into multiple sources (including Kafka, via the dedicated plugin). I'm sure there are other similar solutions.
There are a couple of important benefits that I see if you use a dedicated solution, instead of building it yourself:
The level of configurability, which is definitely useful in the future (in a dedicated solution, you need to write code each time you want to build something new)
The number of destinations where you can export the logs
Support for a hybrid architecture - with a few config updates, you will be able to grab logs from non-JVM microservices
I am using vert.x as an api gateway to route calls to downstream services.
As of now, I am using single web client instance which is shared across multiple verticles (injected through guice)
Does it make sense for each verticle to have it's own webclient? Will it help in boosting performance? (My each gateway instance runs 64 vericles and handles approximately 1000 requests per second)
What are the pros and cons of each approach?
Can someone help to figure out what's the ideal strategy for the same?
Thanks
Vert.x is optimized for using a single WebClient per-Verticle. Sharing a single WebClient instance between threads might work, but it can negatively affect performance, and could lead to some code running on the "wrong" event-loop thread, as described by Julien Viet, the lead developer of Vert.x:
So if you share a web client between verticles, then your verticle
might reuse a connection previously open (because of pooling) and you
will get callbacks on the event loop you won't expect. In addition
there is synchronization in the web client that might become contented
when used intensively from different threads.
Additionally, the Vert.x documentation for HttpClient, which is the underlying object used by WebClient, explicitly states not to share it between Vert.x Contexts (each Verticle gets its own Context):
The HttpClient can be used in a Verticle or embedded.
When used in a Verticle, the Verticle should use its own client
instance.
More generally a client should not be shared between different Vert.x
contexts as it can lead to unexpected behavior.
For example a keep-alive connection will call the client handlers on
the context of the request that opened the connection, subsequent
requests will use the same context.
Here is the background:
We have a cluster (of 3) different services deployed on various containers (like Tomcat, TomEE, JBoss) etc. Each of the services does one thing. Like one service manages a common DB and provides REST services to CRUD the db. One service puts some data into a JMS Queue, Another service reads from the Queue and updates the DB. There is a client app that makes a REST service call to one of the service that sets off creating a row in the db, pushing that row into a queue etc.
Question: We need to implement the client app so that we know at any given point in time where the processing is. How do I implement this in RcJava 2/Java 9?
First, you need to determine what functionality in RxJava 2 will benefit you.
Coordination between asynchronous sources. Since you have a) event-driven requests from one side, and b) network queries on the other sides, this is a good fit so far.
Managing a stream of data, transforming and combining from one or more sources. You have given no indication that this is required.
Second, you need to determine what RxJava 2 does not provide:
Network connections. This is provided by your existing libraries.
Database management. Again, this is provided in your existing solutions.
Now, you have to decide whether the firstlies add up to something you can benefit from, given the up-front costs of learning a new library.
I have an orchestrator service which keeps track of the instances that are running and what request they are currently dealing with. If a new instance is required, I make a REST call to increase the instances and wait for the new instance to connect to the orchestrator. It's one request per instance.
The orchestrator tracks whether an instance is doing anything and knows which instances can be stopped, however there is nothing in the API that allows me to reduce the number of instances stopping a particular instance, which is what I am trying to achieve.
Is there anything I can do to manipulate the platform into deterministically stopping the instances that I want to stop? Perhaps by having long running HTTP requests to the instances I require and killing the request when it's no longer required, then making the API call to reduce the number of instances?
Part of the issue here is that I don't know the specifics of the current behavior...
Assuming you're talking about CloudFoundry/Instant Runtime applications, all of the instances of an applications are running behind a load balancer which uses round-robin to distribute requests across the instances (unless you have session affinity cookie set up). Differentiating between each instances for incoming requests or manual scaling is not recommended and it's an anti-pattern. You can not control which instance the scale down task will choose.
If you really want that level of control with each instance, maybe you should deploy them as separate applications. MyApp1, MyApp2, MyApp3, etc. All of your applications can have the same route (myapp.mybluemix.net). Each of the applications can now distinguish themselves by their name (VCAP_APPLICATION) allowing you terminate them.
I have an Azure hosted web application which works alongside a number of instances of a worker role. Currently the web app passes work to these workers by placing messages in an Azure queue for the workers to pick up. The workers pass status and progress messages back by placing messages into a 'feedback' queue. At the moment, in order to inform my browser clients as to progress, I make ajax based periodic polling calls in the browser to an MVC controller method which in turn reads the Azure 'feedback' queue and returns these messages as json back to the browser.
Obviously, SignalR looks like a very attractive alternative to this clumsy polling / queing approach, but I have found very little guidance on how to go about doing this when we are talking about multiple worker roles (as opposed to the web role) needing to send status to individual or all clients .
The SignalR.WindowsAzureServiceBus by Clemens vasters looks superb but leaves one a bit high and dry at the end i.e. a good example solution is lacking.
Added commentary: From my reading so far it seems that no direct communication from worker role (as opposed to web role) to browser client via the SignalR approach is possible. It seems that workers have to communicate with the web role using queues. This in turn forces a polling approach ie the queues must be polled for messages from the worker roles - this polling has to originate (be driven from) from the browser it appears (how can a polling loop be set up in a web role?)
In summary, SignalR, even with the SignalR.WindowsAzureServiceBus scale out approach of Clemens Vasters, cannot handle direct comunication from worker role to browser.
Any comments from the experts would be appreciated.
You can use your worker roles as SignalR clients, so they will send messages to the web role (which is SignalR server) and the web role in turn will forward messages to clients.
We use Azure Service Bus Queues to send data to our SignalR web roles which then forward on to the clients.
The CAT pages have very good examples of how to set up asynchronous loops and sending.
Keep in mind my knowledge of this two technologies is very basic, I'm just starting. And I might have misunderstood your question, but it seems pretty obvious to me:
Web roles are capable of subscribing to a queue server where the worker role deposits the message?. If so there would be no client "pulling", the queue service would provide the web server side code with a new message, and through SignalR you would push changes to the client without client requests involved. The communication between web and worker would remain the same (which in my opinion, it's the proper way to do it).
If you are using the one of the SignalR scaleout backplanes you can get workers talking to connected clients via your web application.
How to publish messages using the SignalR SqlMessageBus explains how to do this.
It also links to a fully worked example which demonstrates one way of doing this.
Alternative message bus products such as NServiceBus could be worth investigating. NServiceBus has the ability to deliver messages asynchronously across process boundaries without the need for polling.