The FluentMigrator documentation recommends running migrations in-process upon application start, but I don't see any commentary about whether this is safe or recommended when the application is running on multiple servers.
I have always gone the paranoid route of using an out-of-process runner before deploying my applications by triggering it from my build servers. Is that paranoia warranted? Will the Transaction-Per-Migration defaults be enough to guarantee that no migration conflicts can occur on the same codebase trying to run simultaneous upgrade migrations?
At present, this functionality is not yet supported and the documentation has been updated to recommend an out-of-process migration runner when your application is run from multiple processes.
There is currently an issue on Github requesting this feature as well as a workaround specifically for SQL Server 2008+ using sp_getapplock.
Related
I have a dockerized ASP.NET web api that I am running on AWS. I am planning on using RDS for the database, and I need to run migrations and I am unsure how I should go about this. My docker container only contains the dotnet runtime, so I can't just SSH into one of the machines and migrate. The RDS instance is set to only accept traffic from within the VPC, so I can't just run them from my machine. What would be the best way to run EF Core migrations into RDS?
I was thinking of maybe setting up a temporary EC2 machine, installing the dotnet SDK, EF Core and the source code, then running migrations and tearing it down. But I don't know if this is a good idea, or whether there is a better way.
A temporary EC2 instance for performing this sort of thing is fine, and a common practice.
I would suggest an alternative of building an AWS CodeBuild job to perform the migration task. However you might find your temporary EC2 instance useful for other things, like connecting to the database to perform ad hoc queries.
I have an web application that runs in JBoss and i use FlyWay in command line mode to migrate between database versions.
I want integrate Flyway in to my application using Spring Framework and only stop the deployment in to the web server if the db version is wrong. Is this possible?
First of all, I strongly recommend migrating the database on application startup so a mismatch can never occur.
If this is not an option, you can implement custom logic to check the results of Flyway.info() against the desired version.
Say your app gets hit with enough users to grind to a halt, does it fire up another instance? How do you plan to get around this in the future?
Ian
Not sure if you mean apps built with 'meteor bundle' or apps deployed on the free hosting with 'meteor deploy'.
Apps deployed to the hosted servers with 'meteor deploy' do not yet have any guarantees or SLAs about scaling. However, the servers can handle a fair bit of load. meteor.com is hosted on meteor deploy, and it survived the Meteor launch.
A server bundle generated with 'meteor bundle' is basically a single process app. It is up to you wire it up to multiple instances, or however you want to implement auto-scaling.
I'm precisely not clear with your question. aren't asking about the meteor framework capability for scalability or meteor app hosting environment scalability. never mind, here it goes
As Meteor framework ecosystem grows, your question of scalability will be answered, currently you can invoke as many as small modules of code through Packages.
Meteor has something called meteor bundle for more which eventually let deploy your application any where around the globe, then you can have your own infrastructure set-up to run your app (something like a network load balancer hosted NodeJS servers and scalable multiple MongoDB instance)
In general, Meteor can scale using by bundling and use your own infrastructure, like mongo, application server. Like #n1mmy said, when the application is bundled, you can specify a certain mongo end point from which you can scale the data storage. In addition, when the bundle is running, it is actually running on nodejs, which means you can scale your web application like nodejs apps. I think in both way can ease your problem if your application has too many users.
Theoretically Meteor scales just fine, but if it really works for you in practise all depends on how you handle your data. This is an excellent post on scalability in Meteor:
https://www.discovermeteor.com/blog/scaling-meteor-the-challenges-of-realtime-apps/
I currently work with a legacy asp.net web application and one of the requirements going forward is that it be deployable to windows azure.
I would like to know how difficult it will be to manage deployment to both Azure and a traditional IIS web server.
Azure seems to require a specific customized version of a web applicaiton project is it possible to deploy the customized web application to a standard IIS instance once it has been converted.
EDIT:
It is a ASP.NET Web Application rather than a Web Site (compiles everything into one dll)
UPDATE:
In the end due to the amount of work involved in converting the application to work in Azure and the cost of Azure compared with other cloud solutions it was decided to go with a traditional Cloud hosted virtual server.
And thank you for the really good answers.
Whether or not you can deploy your application to Azure almost as is depends a lot on how your application works. Azure pretty much requires your application be stateless. If it's a plain vanilla web application that stores data in the session or application cache only and saves data to a database only, then you can deploy it to Azure.
If you have stateful services running like background threads (which is bad anyways), or if you save data to the file system (besides temporary caching), then you may have issues. Really, the issues moving to Azure are really the same as moving to any multi-server load balanced solution. One caveat is permanent storage.
If you need to store data in a place other than the database, then you're best off working with Azure's storage solution which has an API and client library for storing binary data, key/value data (they call it tables, but really, it's not tables), and queues. They also do have a transparent blob-as-file-system option for compatibility. If you want to use these in your app that also is used outside of Azure then you need to write an extra layer between your code and the Azure client library that supports both Azure services and standard local service. Azure SDK does include emulators for Azure services, but they're definitely not meant for production use.
As far as the mechanics of Azure-specific projects, that is actually not that difficult. Yes, you need to create an Azure-specific project in your solution that defines the Web Role and what gets deployed, but it will reference your existing Web Application, not the other way around. You can deploy the Azure Web Role to Azure or you can continue to deploy the existing application to IIS normally and concurrently.
Web Site, Web Application, MVC, really doesn't make much of a difference. Actually doesn't have to be .NET either. Can be PHP or Java or whatever you want to put on your VM. It'll all work the same as far as Azure is concerned.
MS likes to push Azure as a Platform-as-a-Service (Paas) solution where they have a ton of services they offer and you run apps on their standard platform, and contrasts that with Amazon AWS which they call Infrastructure-as-a-Service (Iaas) which is "just" a Virtual Machine. However, MS is really just as much a IaaS solution as AWS, perhaps even more so. The only difference between AWS and Azure is AWS allows you to choose what to install on your VM and with Azure you have to use Windows Server 2008 R2 as the basis for your VM (but you can customize the VM image to install custom software on top of windows). With both Azure and AWS, the hosts offer additional PaaS services you can take advantage of for data storage and message routing. AWS also offers tons of extra services like video streaming.
Also note that with Azure (and AWS I think) you can use the services they offer even in a non-hosted application. If you want to use Azure's data storage from a non-Azure application, you can do that, it's just HTTP REST calls to get/put data. The only differences you pay for data in/out between datacenter and your non-datacenter-hosted application which would be free if the app was also inside the datacenter (just the data in/out is free in-datacenter, you still have storage and transaction fees).
A few things:
Samuel Neff's answer mentioned mounting a file system in a blob (a Cloud Drive). Only one instance may lock this cloud drive for writing, so it does not behave like a network file share. You'll need to plan for this.
You'll need to integrate with the Windows Azure diagnostics subsystem, to gain visibility into your app's run state (e.g. performance counters, trace logs, etc.).
If there are 3rd-party apps that your web app depends on, you'll need to install these. These actually get installed as part of the role instance's boot process, either via your OnStart() event handler or as a startup task. The latter allows for admin-level installs (including registry changes, COM component installations, etc.). You'll need to carefully manage these installations, as they impact the boot time of the instance.
For an asp.net app, you'll need to think about session state. In-proc session state won't work, because each instance will have its own state store in memory. The SQL Azure session state provider doesn't have background cleanup agents, so you'll need to build this into your web or worker role instance (see this blog post by the SQL Azure team for the implementation). The best option is to use the AppFabric Cache, a new service that just went into production. This cache-as-a-service provides an custom session state provider for asp.net as well. Note: As of today, the AppFabric Cache service is only accessible via a .NET interface; there's no REST interface for it (all other storage services - tables, blobs, queues - have a REST interface). .NET, Java, and PHP all have storage client libraries. Ruby has one from the open source community.
You'll have to manage scaling out to more than one instance, when the need arises. This is not a built-in service today, but there are 3rd-party services such as ParaLeap's AzureWatch. There's also Microsoft's System Center Operations Manager, which now has Windows Azure monitoring support. You'll also need to handle scale-back situations, where you reduce the number of server instances.
I have some additional details in an answer for a similar StackOverflow question, here.
I have not tried Windows Azure Migration Scanner personally, but if it works as advertised, this would really come in handy.
Usually I would look at writing a Windows Service to manage tasks that aren't suited to being hosted in a web application. These types of tasks are usually long running processes or scheduled tasks. Although this is normally the primary approach for these types of tasks, people have looked at ways of running these kinds of background processes in a web application by kicking off a number of threads in the Application_Start event exposed by Global.asax. The problem with this approach has always been that if your IIS worker process dies, then your background thread is killed too (effectively your 'Windows Service' is stopped until the next request is received).
ASP .NET 4.0 offers a solution to this problem. You can now set the StartMode to 'AlwaysRunning' as described in this blog post by Scott Gu. Somewhere in the comments on this post, someone asks a question about the viability of hosting Windows Service type tasks in IIS since the new feature ensures the worker process is always running. Scott mentioned that it would definitely support the scenario. Further to this, the recent introduction of AppFabric means that Microsoft themselves are providing simple hooks for hosting and monitoring WCF and WF services in a web application.
What does this mean for those of us that used to write Windows Services to support our web apps? Should we adopt this model? What are the pitfalls? As far as I can tell, there are a number of benefits to hosting 'Windows Service' processes in a web application, the most useful being the ease of deployment. Furthermore, we can actually start developing simple user interfaces to our services which provide information about what is happening at runtime.
If I had to go this route, I don't think that I would host my 'Windows Service' type functionality in the customer facing web application. I would probably develop a new web application project (much like I would in the Windows Service context) that would host my long running/scheduled task processes. I guess there are few reasons for this.
Security. There may be a different security model for the UI displaying information about the running background processes. I would not want to expose this UI to anyone else but the ops team. Also, the web application may run as a different user which has an elevated set of permissions.
Maintenance. It would be great to be able to deploy changes to the application hosting the background processes without impacting on user's using the front end website.
Performance. Having the application separated from the main site processing user requests means that background threads will not diminish IIS's capability to handle the incoming request queue. Furthermore, the application processing the background tasks could be deployed to a separate server if required.
I would be really interested to hear your thoughts on this approach and whether I should be sticking with Windows Services. I am very tempted to try this new approach.
What does this mean for those of us that used to write Windows Services to support our web apps?
I think this a key scenario where you could be move away from a Windows service to using the continous running web site.
Should we adopt this model?
Standard development answer: Depends ;)
What are the pitfalls?
One issue I can see is the IIS dependency. If you need a service to run on a users machine I would not feel comfortable about asking them to install IIS just to run my service. Here I think the traditional model works better.
Monitoring and tracking are major issues, but as you also point out this is solved by AppFabric. It is even better than what you get from the Window Service. However you have added another dependency which also will require .NET 4.0 and a relatively new version of Windows. I could also be wrong here, but my understanding is that AppFabric is not supported in production on client OS's. Which could bring in additional headaches.
You will lose pause functionality in the continuous web site model too.
Finally IIS killing inactive app-pools isn't the only way an app pool can recycle. Editing a web.config file causes it for instance, which may not be an ideal situation.
the most useful being the ease of deployment.
I also think development is much easier - in the past I have had a console app and a windows service so I can dev/test on my machine using the console app and then change it to a windows service when it goes out. Now dev/test is MUCH easier.
A must read for this is Death to Windows Services...Long Live AppFabric!
What are the pitfalls?
One I found, no shutdown event. You have AppStart when the web site starts (not global.asax because that is HTTP only) but you have no way to handle shutdown which could mean disposing becomes an issue.
I would suggest sticking with a windows service. The issue is with your number 2.
You won't be able to update service part of web site without restarting whole web site.