Are AMRMClientAsync and AMNMClientAsync thread-safe? - hadoop-yarn

I'm writing yarn applications and I wonder if these interfaces are thread-safe. When I look up their api, I cannot find any message of it.

After reading the source code, I have found that its api is thread-safe since it has synchronized symbol.

Related

How do you use MTLSharedTextureHandle or MTLSharedEventHandle with the C XPC interface on macOS?

TL;DR: How do you encode and decode an MTLSharedTextureHandle and MTLSharedEventHandler such that it can be transported across an XPC connection inside an xpc_dictionary?
A macOS application I'm working on makes extensive use of XPC services and was implemented using the C-based API. (i.e.: xpc_main, xpc_connection, xpc_dictionary...) This made sense at the time because certain objects, like IOSurfaces, did not support NSCoding/NSSecureCoding and had to be passed using IOSurfaceCreateXPCObject.
In macOS 10.14, Apple introduced new classes for sharing Metal textures and events between processes: MTLSharedTextureHandle and MTLSharedEventHandle. These classes support NSSecureCoding but they don't appear to have a counter-part in the C-XPC interface for encoding/decoding them.
I thought I could use something like [NSKeyedArchiver archivedDataWithRootObject:requiringSecureCoding:error] to just convert them to NSData objects, which can then be stored in an xpc_dictionary, but when I try and do that, I get the following exception:
Caught exception during archival:
This object may only be encoded by an NSXPCCoder.
(NSXPCCoder is a private class.)
This happens for both MTLSharedTextureHandle and MTLSharedEventHandle. I could switch over to using the new NSXPCConnection API but I've already got an extensive amount of code built on the C-interface, so I'd rather not have to make the switch.
Is there any way to archive either of those two classes into a payload that can be stored in an xpc_dictionary for transfer between the service and the client?
MTLSharedTextureHandle only works with NSXPCConnection. If you're creating the texture from an IOSurface you can share the surface instead which is effectively the same thing. Make sure you are using the same GPU (same id<MTLDevice>) in both processes.
There is no workaround for MTLSharedEventHandle using public API.
I recommend switching to NSXPCConnection if you can. Unfortunately there isn't a good story for partially changing over using public API, you'll have to do it all at once or split your XPC service into two separate services.

Profiler lib for wcf + postsharp

We need to add a new profiling feature to our WCF application, for logging where time is spendt in the application. I'm looking at PostSharp for a convention driven approach of applying the logging and need some input on how to actually log it. I've already created a custom class for logging purposes, using StopWatch and can log the steps through the layers of my WCF application. However I'm wondering if there's a thread safe alternative library I could use in conjunction with PostSharp for this purpose. I've come across MiniProfiler, but it seems to be intended for ASP.NET MVC applications mainly. Any other frameworks I should consider or should I just use my custom class?
Thanks
I did something like that in the past using a IClientMessageInspector implemented on a custom IEndpointBehavior.
Depending on what kind of logging you want, this might just do the trick. There's an example in the following link
IClientMessageInspector Interface
PostSharp itself is thread-safe. The aspects that you write may be thread-unsafe if poorly written, but there's always a way to make them thread-safe.
If you're using OnMethodBoundaryAspect and need to pass something from OnEntry to OnSuccess, store the initial stopwatch value in OnMethodExecutionArgs.MethodExecutionTag.

Creating an Objective-C API

I have never made an API in objective-c, and need to do this now.
The "idea" is that I build an API which can be implemented into other applications. Much like Flurry, only for other purposes.
When starting the API, an username, password and mode should be entered. The mode should either be LIVE or BETA (I guess this should be an NSString(?)), then afterwards is should be fine with [MyAPI doSomething:withThisObject]; ect.
So to start it [MyAPI username:#"Username" password:#"Password" mode:#"BETA"];
Can anyone help me out with some tutorials and pointer on how to learn this best?
It sounds like what you want to do is build a static library. This is a compiled .a file containing object code that you'll distribute to a client along with a header file containing the interface. This post is a little outdated but has some good starting points. Or, if you don't mind giving away your source code, you could just deliver a collection of source files to your client.
In terms of developing the API itself, it should be very similar to the way you'd design interfaces and implementations of Objective-C objects in your own apps. You'll have a MyAPI class with functions for initialization, destruction, and all the functionality you want. You could also have multiple classes with different functionality if the interface is complex. Because you've capitalized MyAPI in your code snippet, it looks like you want to use it by calling the class rather than an instance of the class - which is a great strategy if you think you'll only ever need one instance. To accomplish this you can use the singleton pattern.
Because you've used a username and password, I imagine your API will interface with the web internally. I've found parsing JSON to be very straightforward in Objective-C - it's easy to send requests and get information from a server.
Personally I would use an enum of unsigned ints rather than a NSString just because it simplifies comparisons and such. So you could do something like:
enum {
MYAPI_MODE_BETA,
MYAPI_MODE_LIVE,
NUM_MYAPI_MODES
};
And then call:
[MyAPI username:#"Username" password:#"Password" mode:MYAPI_MODE_BETA];
Also makes it easy to check if they've supplied a valid mode. (Must be less than NUM_MYAPI_MODES.)
Good luck!

DLL Reflection?

Is something like this possible? If so, could you point me in the right direction for learning how?
applicationx tries to run the method start() in dll_one.dll
dll_one.dll runs the command
applicationx tries to run the method run() in dll_one.dll
dll_one.dll doesn't have a method run() and hasn't prepared for such an occurance.
dll_one.dll asks dll_two.dll if it has a run()
dll_two runs run()
Basically, I want it so if dllA doesn't have a method that the application is looking for, it asks dllB. This is assuming, as well, that ApplicationX and dllB don't know anything about dllA and dllA kind of just appeared out of nowhere (I want dlls dynamically like a patch to my applications without having to rewrite ALL of the methods, properties, etc. in the dll and have everything else just routed to the old dll).
Any ideas? Keep in mind, I'm using vb.net so a .net reference is appreciated.
It seems like you're asking for a plug-in architecture for your app (except that "patch" part is bothering me). If so, you can try MEF, which solves this exact problem.
The specific thing you ask for isn't possible. You can't have a non-existent method call automatically re-routed to a different dll. You can't "run the method run() in dll_one.dll" unless you've compiled that code, and it won't compile if the method doesn't exist. You also can't compile code against dllB and then drop dllA in and have it intercept method calls. Reflection could conceivably solve part of your problem, but you'd not want to base your code around calling all methods by reflection - it'd be horrendously unperformant and not very maintainable.
As Anton suggests, a plugin approach might work. However, this would rely on you being able to specify up-front the interface for your plugin, which sounds like it would contradict your original requirement.
Another problem: if you'd not deployed dllA until later, how would your ApplicationX know to call method start() in dll_one.dll anyway? You'd surely need to re-deploy at least the base application for that part to work.
These kinds of problem are often best solved by having a more specific set of requirements to work to: what functionality are you likely to want to extend or change in the future? Could you support a common set of interfaces that allow extensibility via plugins, or can you need to redeploy encapsulated chunks of your application with new functionality? Is there UI involved or is this just to change back-end logic? Questions like this could help to suggest more viable solutions.

Sending interfaces as message in NServiceBus with the Binary Serializer

I recently moved to using the binary serializer to send messages with NServiceBus. My messages are all defined as interfaces and are instantiated using
bus.Send<MessageType>(msg => msg.Property = someValue)
This leads to an exception being thrown from NServiceBus stating that
Cannot create an instance of an
interface
I can see from the stack trace that the SimpleMessageMapper is being used, and after looking in the source can see it's making a call to Activator.CreateInstance.
I can't find anything in the documentation stating that it's not possible to do what I'm trying to do, is there a way to fix this?
Thanks,
Matt
I only just started playing with nServiceBus, so all I can offer you is theory :).
Are you defining the implementation classes for your message interfaces, or is nServiceBus generating classes on its own? If the former, make sure you still have a default constructor and that the class and all fields/events are marked as [Serializable] or [NonSerialized]. If the latter, it's possible that nServiceBus doesn't know how to generate members which may be needed for (de)serialization. You may have to write and map the implementation class yourself.