I have some code that needs to run as a result of a call to a service bus. This particular code is CPU intensive and it is possible that 100s of these will need to run at the same time. Does Azure Web Jobs use computing resources from one machine, or does it use any available computing resources from several machines?
Web Jobs uses your web app resources, why don't you try the Azure Functions which can be scaled and their pricing is almost zero. They are in Technical preview i have tried using it, Azure functions is very cheap. If your service bus call can be out of process service meaning it does not need a instance results from your application you can try azure functions. Azure functions are mostly used for maintenance and night time running jobs. I have used it to minify my images to thumbnail. It worked perfectly fine
Azure Webjobs is designed to run on as many servers as you've scaled up the website to run on. By default it will run up to 16 tasks from a queue concurrently but this is configurable as shown here.
public class Program
{
static void Main()
{
JobHostConfiguration config = new JobHostConfiguration();
config.Queues.BatchSize= 1;
JobHost host = new JobHost(config);
host.RunAndBlock();
}
}
A web job deployed as part of an Azure WebSite will share the resources with the web application. You have the option to scale the website up to 10 (I think?) instances if you need parallelism.
As I mentioned in the comment to Matthew's post, if you use the Azure WebJobs SDK to have functions triggered by ServiceBus queue messages, we don't parallelize as part of the same host. This means that messages will be processed sequentially as long as you have a single host.
Related
I'm developing an application with microservice architecture running on Google Cloud Run (fully managed). I want to add communication over events to my services. As I know, the only option is to use Eventarc. I'm curious what is the best way to reproduce the event-driven design when developing locally and how to make deployment as seamless as possible.
Not familiar with Google cloud explicitly, but I assume they all work similarly. As long as you can get your code running locally, then you can still use the cloud hosted message queue / pub/sub interface from your local code.
This way you can debug and try things out on your local machine while still using the messaging / eventing infrastructure.
If I have two distinct WebJobs, both triggered with TimerTrigger and at the same moment in time, can they ever run concurrently on the same host or will one wait for the other to complete?
In other words can a single host run two distinct webjobs in parallel which are triggered via TimerTrigger?
The main reason I am asking is because this project is heavily using both entity framework (DbContext) and also dependency injection and I want to be informed when planning my DI strategy. The solution for me may be challenging if two TimerTrigger webjobs can run concurrently on the same host (because services in this project receive an injected DbContext expecting a unit-of-work lifetime). In this case I might have to synchronize execution of the webjobs (within the same host) myself...
Yes, you can have many webjobs running at the time time. As long as they aren't configured to trigger based on a queue based trigger where an incoming message might cause conflict. You should be fine with a TimerTrigger.
I am trying to understand Serverless architecture which says 2 distinct things:
you as an app developer think about your function only and not about the server responsibilities. Well, the server still has got to be somewhere. By servers, I understand here that its mean both:
on the infrastructure side Physical Server/VM/container
as well as the on the software side: say, Tomcat
Now, I have worked on Cloud Foundry and studied the ER i.e. Diego Architecture of Cloud Foundry and the buildpack and open Service Broker API facility of Cloud foundry. Effectively, Cloud Foundry also already works on a "similar" model where the application developer focuses on his code and the deployment model with the help of buildpack prepares a droplet with the needed Java runtime and Tomcat runtime and then uses it to create a garden container that serves user requests. So, the developer does not have to worry about where the Tomcat server or the VM/container will come from. So, aren't we already meeting that mandate in Cloud Foundry?
your code comes into existence for the duration of execution and then dies. This I agree is different from the apps/microserevices that we write in Cloud Foundry in that they are long running server processes instead. Now, if I were to develop a Java webapp/microservice with 3 REST endpoints (myapp/resource1, myapp/resource2, myapp/resource3) possibly on a Tomcat Web Server, I need:
a physical machine or a VM or a container,
the Java runtime
the Tomcat container to be able to run my war file.
Going by what Serverless suggests, I infer I am supposed to concentrate only on the very specific function say handling the request to myapp/resource1. Now, in such a scenario:
What is my corresponding Java class supposed to look like?
Where do I get access to the J2EE objects like HttpServletRequest or HttpServletResponse objects and other http or servlet or JAX-RS or Spring MVC provided objects that are created by the Tomcat runtime?
Is my Java class executed within a container that is created for the duration of execution and then destroyed after execution? If yes, who manages the creation/destruction of such a container?
Would Tomcat even be required? Is there an altogether different generic way of handling requests to these three REST endpoints? Is it somewhat like httpd servers using python/Java CGI scripts to handle http requests?
We have developed a custom JAX-WS application that essentially achieves two things.
Exposes a few web service methods to perform some functionality.
Utilizes org.quartz.Scheduler to schedule and execute some polling tasks that monitors and processes data on a few database tables. (The logic here is slightly complex, hence a custom application was chosen over the use of WSO2 DSS)
This application is uploaded on WSO2 AS 5.2.1 and runs quite seamlessly. However, I'm unsure what will happen if we have to cluster the AS application server. Logically, I would think that each node will have its own instance of the custom application running within it, and hence its own scheduler. Would this not increase the risk of processing the same record, across both instances. Is my interpretation of the above scenario correct, from a clustering perspective?
Yes.You are correct.In cluster of app server nodes each nodes will have its own instance of the application.In your case each node will have seperate scheduler.You may consider using tasks from ESB 4.9.0. there WSO2 has added coordination support to work in cluster environment.
I am building a web application, not hosted by the client but on the cloud somewhere, probably azure. It will need the ability to access some local devices such as a receipt printer or cash draw from where they run it. Via a browser this is obviously not possible.
I was thinking of creating a service that they can install locally on their PC that listens to a web service and when it gets a message can perform the tasks such as opening the cash draw.
Is this how people have done similar things in the past? Are there other methods? What technologies do you recommend as I plan on the local service being able to be run on either PC or Mac with minimum fuss. I was considering either Flex or Mono (the app was written in ASP.NET), any other suggestions?