How should I stream the output from one program to an undefined number of programs in such a fashion that the data isn't buffered anywhere and that the application where the stream originates from doesn't block even if there's nothing reading the stream, but the programs reading the stream do block if there's no output from the first-mentioned program?
I've been trying to Google around for a while now, but all I find is methods where the program does block if nothing is reading the stream.
How should I stream the output from one program to an undefined number of programs in such a fashion that the data isn't buffered anywhere and that the application where the stream originates from doesn't block even if there's nothing reading the stream
Your requirements as stated can not possibly be satisfied without some form of a buffer.
Most straightforward option is to write the output to the file and let consumers read that file.
Another option is to have a ring-buffer in a form of a memory mapped file. As the capacity of a ring-buffer is normally fixed there needs to be a policy for dealing with slow consumers. Options are: block the producer; terminate the slow consumer; let the slow consumer somehow recover when it missed data.
Many years ago I wrote something like what you describe for an audio stream processing app (http://hewgill.com/nwr/). It's on github as splitter.cpp and has a small man page.
The splitter program currently does not support dynamically changing the set of output programs. The output programs are fixed when the command is started.
Without knowing exactly what sort of data you are talking about (how large is the data, what format is it, etc, etc) it is hard to come up with a concrete answer. Let's say for example you want a "ticker-tape" application that sends out information for share purchases on the stock exchange, you could quite easily have a server that accepts a socket from each application, starts a thread and sends the relevant data as it appears from the recoder at the stock market. I'm not aware of any "multiplexer" that exists today (but Greg's one may be a starting point). If you use (for example) XML to package the data, you could send the second half of a packet, and the client code would detect that it's not complete, so throws it away.
If, on the other hand, you are sending out high detail live update weather maps for the whole country, the data is probably large enough that you don't want to wait for a full new one to arrive, so you need some sort of lock'n'load protocol that sets the current updated map, and then sends that one out until (say) 1 minute later you have a new one. Again, it's not that complex to write some code to do this, but it's quite a different set of code to the "ticker tape" solution above, because the packet of data is larger, and getting "half a packet" is quite wasteful and completely useless.
If you are streaming live video from the 2016 Olympics in Brazil, then you probably want a further diffferent solution, as timing is everything with video, and you need the client to buffer, pick up key-frames, throw away "stale" frames, etc, etc, and the server will have to be different.
Related
I'm working on what's basically a highly-available distributed message-passing system. The system receives messages from someplace over HTTP or TCP, perform various transformations on it, and then sends it to one or more destinations (also using TCP/HTTP).
The system has a requirement that all messages sent to a given destination are in-order, because some messages build on the content of previous ones. This limits us to processing the messages sequentially, which takes about 750ms per message. So if someone sends us, for example, one message every 250ms, we're forced to queue the messages behind each other. This eventually introduces intolerable delay in message processing under high load, as each message may have to wait for hundreds of other messages to be processed before it gets its turn.
In order to solve this problem, I want to be able to parallelize our message processing without breaking the requirement that we send them in-order.
We can easily scale our processing horizontally. The missing piece is a way to ensure that, even if messages are processed out-of-order, they are "resequenced" and sent to the destinations in the order in which they were received. I'm trying to find the best way to achieve that.
Apache Camel has a thing called a Resequencer that does this, and it includes a nice diagram (which I don't have enough rep to embed directly). This is exactly what I want: something that takes out-of-order messages and puts them in-order.
But, I don't want it to be written in Java, and I need the solution to be highly available (i.e. resistant to typical system failures like crashes or system restarts) which I don't think Apache Camel offers.
Our application is written in Node.js, with Redis and Postgresql for data persistence. We use the Kue library for our message queues. Although Kue offers priority queueing, the featureset is too limited for the use-case described above, so I think we need an alternative technology to work in tandem with Kue to resequence our messages.
I was trying to research this topic online, and I can't find as much information as I expected. It seems like the type of distributed architecture pattern that would have articles and implementations galore, but I don't see that many. Searching for things like "message resequencing", "out of order processing", "parallelizing message processing", etc. turn up solutions that mostly just relax the "in-order" requirements based on partitions or topics or whatnot. Alternatively, they talk about parallelization on a single machine. I need a solution that:
Can handle processing on multiple messages simultaneously in any order.
Will always send messages in the order in which they arrived in the system, no matter what order they were processed in.
Is usable from Node.js
Can operate in a HA environment (i.e. multiple instances of it running on the same message queue at once w/o inconsistencies.)
Our current plan, which makes sense to me but which I cannot find described anywhere online, is to use Redis to maintain sets of in-progress and ready-to-send messages, sorted by their arrival time. Roughly, it works like this:
When a message is received, that message is put on the in-progress set.
When message processing is finished, that message is put on the ready-to-send set.
Whenever there's the same message at the front of both the in-progress and ready-to-send sets, that message can be sent and it will be in order.
I would write a small Node library that implements this behavior with a priority-queue-esque API using atomic Redis transactions. But this is just something I came up with myself, so I am wondering: Are there other technologies (ideally using the Node/Redis stack we're already on) that are out there for solving the problem of resequencing out-of-order messages? Or is there some other term for this problem that I can use as a keyword for research? Thanks for your help!
This is a common problem, so there are surely many solutions available. This is also quite a simple problem, and a good learning opportunity in the field of distributed systems. I would suggest writing your own.
You're going to have a few problems building this, namely
2: Exactly-once delivery
1: Guaranteed order of messages
2: Exactly-once delivery
You've found number 1, and you're solving this by resequencing them in redis, which is an ok solution. The other one, however, is not solved.
It looks like your architecture is not geared towards fault tolerance, so currently, if a server craches, you restart it and continue with your life. This works fine when processing all requests sequentially, because then you know exactly when you crashed, based on what the last successfully completed request was.
What you need is either a strategy for finding out what requests you actually completed, and which ones failed, or a well-written apology letter to send to your customers when something crashes.
If Redis is not sharded, it is strongly consistent. It will fail and possibly lose all data if that single node crashes, but you will not have any problems with out-of-order data, or data popping in and out of existance. A single Redis node can thus hold the guarantee that if a message is inserted into the to-process-set, and then into the done-set, no node will see the message in the done-set without it also being in the to-process-set.
How I would do it
Using redis seems like too much fuzz, assuming that the messages are not huge, and that losing them is ok if a process crashes, and that running them more than once, or even multiple copies of a single request at the same time is not a problem.
I would recommend setting up a supervisor server that takes incoming requests, dispatches each to a randomly chosen slave, stores the responses and puts them back in order again before sending them on. You said you expected the processing to take 750ms. If a slave hasn't responded within say 2 seconds, dispatch it again to another node randomly within 0-1 seconds. The first one responding is the one we're going to use. Beware of duplicate responses.
If the retry request also fails, double the maximum wait time. After 5 failures or so, each waiting up to twice (or any multiple greater than one) as long as the previous one, we probably have a permanent error, so we should probably ask for human intervention. This algorithm is called exponential backoff, and prevents a sudden spike in requests from taking down the entire cluster. Not using a random interval, and retrying after n seconds would probably cause a DOS-attack every n seconds until the cluster dies, if it ever gets a big enough load spike.
There are many ways this could fail, so make sure this system is not the only place data is stored. However, this will probably work 99+% of the time, it's probably at least as good as your current system, and you can implement it in a few hundred lines of code. Just make sure your supervisor is using asynchronous requests so that you can handle retries and timeouts. Javascript is by nature single-threaded, so this is slightly trickier than normal, but I'm confident you can do it.
I've a readable NodeJS Stream which I want to use twice. Disclaimer: I'm not very comfortable with streams
Why?
My Service allows uploading of images for users. I want to avoid uploading of the same images.
My workflow is as follows:
upload image per ajax
get hash of image
if hash in database
return url from database
else
pass hash to resize&optimize pipeline
upload image to s3 bucket
get hash of image and write it to database with url
return s3 url
I get the hash of my stream with hashstream and optimize my image with gm.
Hashstream takes a stream, closes it, creates a hash and returns it with a callback.
My question is: What would be the best approach to combine both methods?
There are two ways to solve it:
Buffer the stream
Since you don't know if your stream will be used again, you can simply buffer it up somehow (somehow meaning handling data events, or using some module, for
example accum). As soon as you know what the outcome of the hash function you'd simply write the whole accumulated buffer into the gm stream.
Use stream.pipe twice to "tee"
You probably know the posix command tee, likewise you can push all the data into two places. Here's an example implementation of a tee method in my "scramjet" stream, but I guess for you it'd be quite sufficient to simply pipe twice. Then as soon as you get your hash calculated and run into the first condition I'd simply send an end.
The right choice depends on if you want to conserve memory or CPU. For less memory use two pipes (your optimization process will start, but you'll cancel it before it would output anything). For less CPU and less processes usage I'd go for buffering.
All in all I would consider buffering only if you can easily scale to more incoming images or you know exactly how much load there is and you can handle it. Either way there will be limits and these limit need to be somehow handled, if you can start couple more instances then you should be better of with using more CPU and keeping the memory at a sensible level.
I'm designing a large-scale project, and I think I see a way I could drastically improve performance by taking advantage of multiple cores. However, I have zero experience with multiprocessing, and I'm a little concerned that my ideas might not be good ones.
Idea
The program is a video game that procedurally generates massive amounts of content. Since there's far too much to generate all at once, the program instead tries to generate what it needs as or slightly before it needs it, and expends a large amount of effort trying to predict what it will need in the near future and how near that future is. The entire program, therefore, is built around a task scheduler, which gets passed function objects with bits of metadata attached to help determine what order they should be processed in and calls them in that order.
Motivation
It seems to be like it ought to be easy to make these functions execute concurrently in their own processes. But looking at the documentation for the multiprocessing modules makes me reconsider- there doesn't seem to be any simple way to share large data structures between threads. I can't help but imagine this is intentional.
Questions
So I suppose the fundamental questions I need to know the answers to are thus:
Is there any practical way to allow multiple threads to access the same list/dict/etc... for both reading and writing at the same time? Can I just launch multiple instances of my star generator, give it access to the dict that holds all the stars, and have new objects appear to just pop into existence in the dict from the perspective of other threads (that is, I wouldn't have to explicitly grab the star from the process that made it; I'd just pull it out of the dict as if the main thread had put it there itself).
If not, is there any practical way to allow multiple threads to read the same data structure at the same time, but feed their resultant data back to a main thread to be rolled into that same data structure safely?
Would this design work even if I ensured that no two concurrent functions tried to access the same data structure at the same time, either for reading or for writing?
Can data structures be inherently shared between processes at all, or do I always explicitly have to send data from one process to another as I would with processes communicating over a TCP stream? I know there are objects that abstract away that sort of thing, but I'm asking if it can be done away with entirely; have the object each thread is looking at actually be the same block of memory.
How flexible are the objects that the modules provide to abstract away the communication between processes? Can I use them as a drop-in replacement for data structures used in existing code and not notice any differences? If I do such a thing, would it cause an unmanageable amount of overhead?
Sorry for my naivete, but I don't have a formal computer science education (at least, not yet) and I've never worked with concurrent systems before. Is the idea I'm trying to implement here even remotely practical, or would any solution that allows me to transparently execute arbitrary functions concurrently cause so much overhead that I'd be better off doing everything in one thread?
Example
For maximum clarity, here's an example of how I imagine the system would work:
The UI module has been instructed by the player to move the view over to a certain area of space. It informs the content management module of this, and asks it to make sure that all of the stars the player can currently click on are fully generated and ready to be clicked on.
The content management module checks and sees that a couple of the stars the UI is saying the player could potentially try to interact with have not, in fact, had the details that would show upon click generated yet. It produces a number of Task objects containing the methods of those stars that, when called, will generate the necessary data. It also adds some metadata to these task objects, assuming (possibly based on further information collected from the UI module) that it will be 0.1 seconds before the player tries to click anything, and that stars whose icons are closest to the cursor have the greatest chance of being clicked on and should therefore be requested for a time slightly sooner than the stars further from the cursor. It then adds these objects to the scheduler queue.
The scheduler quickly sorts its queue by how soon each task needs to be done, then pops the first task object off the queue, makes a new process from the function it contains, and then thinks no more about that process, instead just popping another task off the queue and stuffing it into a process too, then the next one, then the next one...
Meanwhile, the new process executes, stores the data it generates on the star object it is a method of, and terminates when it gets to the return statement.
The UI then registers that the player has indeed clicked on a star now, and looks up the data it needs to display on the star object whose representative sprite has been clicked. If the data is there, it displays it; if it isn't, the UI displays a message asking the player to wait and continues repeatedly trying to access the necessary attributes of the star object until it succeeds.
Even though your problem seems very complicated, there is a very easy solution. You can hide away all the complicated stuff of sharing you objects across processes using a proxy.
The basic idea is that you create some manager that manages all your objects that should be shared across processes. This manager then creates its own process where it waits that some other process instructs it to change the object. But enough said. It looks like this:
import multiprocessing as m
manager = m.Manager()
starsdict = manager.dict()
process = Process(target=yourfunction, args=(starsdict,))
process.run()
The object stored in starsdict is not the real dict. instead it sends all changes and requests, you do with it, to its manager. This is called a "proxy", it has almost exactly the same API as the object it mimics. These proxies are pickleable, so you can pass as arguments to functions in new processes (like shown above) or send them through queues.
You can read more about this in the documentation.
I don't know how proxies react if two processes are accessing them simultaneously. Since they're made for parallelism I guess they should be safe, even though I heard they're not. It would be best if you test this yourself or look for it in the documentation.
I have built a standalone app version of a project that until now was just a VST/audiounit. I am providing audio support via rtaudio.
I would like to add MIDI support using rtmidi but it's not clear to me how to synchronise the audio and MIDI parts.
In VST/audiounit land, I am used to MIDI events that have a timestamp indicating their offset in samples from the start of the audio block.
rtmidi provides a delta time in seconds since the previous event, but I am not sure how I should grab those events and how I can work out their time in relation to the current sample in the audio thread.
How do plugin hosts do this?
I can understand how events can be sample accurate on playback, but it's not clear how they could be sample accurate when using realtime input.
rtaudio gives me a callback function. I will run at a low block size (32 samples). I guess I will pass a pointer to an rtmidi instance as the userdata part of the callback and then call midiin->getMessage( &message ); inside the audio callback, but I am not sure if this is thread-sensible.
Many thanks for any tips you can give me
In your case, you don't need to worry about it. Your program should send the MIDI events to the plugin with a timestamp of zero as soon as they arrive. I think you have perhaps misunderstood the idea behind what it means to be "sample accurate".
As #Brad noted in his comment to your question, MIDI is indeed very slow. But that's only part of the problem... when you are working in a block-based environment, incoming MIDI events cannot be processed by the plugin until the start of a block. When computers were slower and block sizes of 512 (or god forbid, >1024) were common, this introduced a non-trivial amount of latency which results in the arrangement not sounding as "tight". Therefore sequencers came up with a clever way to get around this problem. Since the MIDI events are already known ahead of time, these events can be sent to the instrument one block early with an offset in sample frames. The plugin then receives these events at the start of the block, and knows not to start actually processing them until N samples have passed. This is what "sample accurate" means in sequencers.
However, if you are dealing with live input from a keyboard or some sort of other MIDI device, there is no way to "schedule" these events. In fact, by the time you receive them, the clock is already ticking! Therefore these events should just be sent to the plugin at the start of the very next block with an offset of 0. Sequencers such as Ableton Live, which allow a plugin to simultaneously receive both pre-sequenced and live events, simply send any live events with an offset of 0 frames.
Since you are using a very small block size, the worst-case scenario is a latency of .7ms, which isn't too bad at all. In the case of rtmidi, the timestamp does not represent an offset which you need to schedule around, but rather the time which the event was captured. But since you only intend to receive live events (you aren't writing a sequencer, are you?), you can simply pass any incoming MIDI to the plugin right away.
I'm currently working on a personal project: creating a library for realtime audio synthesis in Flash. In short: tools to connect wavegenarators, filters, mixers, etc with eachother and supply the soundcard with raw (realtime) data. Something like max/msp or Reaktor.
I already have some working stuff, but I'm wondering if the basic setup that I wrote is right. I don't want to run into problems later on that force me to change the core of my app (although that can always happen).
Basically, what I do now is start at the end of the chain, at the place where the (raw) sounddata goes 'out' (to the soundcard). To do that, I need to write chunks of bytes (ByteArrays) to an object, and to get that chunk I ask whatever module is connected to my 'Sound Out' module to give me his chunk. That module does the same request to the module that's connected to his input, and that keeps happening until the start of the chain is reached.
Is this the right approach? I can imagine running into problems if there's a feedbackloop, or if there's another module with no output: if i were to connect a spectrumanalyzer somewhere, that would be a dead end in the chain (a module with no outputs, just an input). In my current setup, such a module wouldnt work because i only start calculating from the sound-output module.
Has anyone experience with programming something like this? I'd be very interested in some thoughts about the right approach. (For clarity: i'm not looking for specific Flash-implementations, and that's why i didnt tag this question under flash or actionscript)
I did a similar thing a while back, and I used the same approach as you do - start at the virtual line out, and trace the signal back to the top. I did this per sample though, not per buffer; if I were to write the same application today, I might choose per-buffer instead though, because I suspect it would perform better.
The spectrometer was designed as an insert module, that is, it would only work if both its input and its output were connected, and it would pass its input to the output unchanged.
To handle feedback, I had a special helper module that introduced a 1-sample delay and would only fetch its input once per cycle.
Also, I think doing all your internal processing with floats, and thus arrays of floats as the buffers, would be a lot easier than byte arrays, and it would save you the extra effort of converting between integers and floats all the time.
In later versions you may have different packet rates in different parts of your network.
One example would be if you extend it to transfer data to or from disk. Another example
would be that low data rate control variables such as one controlling echo-delay may, later, become a part of your network. You probably don't want to process control variables with the same frequency that you process audio packets, but they are still 'real time' and part of the function network. They may for example need smoothing to avoid sudden transitions.
As long as you are calling all your functions at the same rate, and all the functions are essentially taking constant-time, your pull-the-data approach will work fine. There will
be little to choose between pulling data and pushing. Pulling is somewhat more natural for playing audio, pushing is somewhat more natural for recording, but either works and ends up making the same calls to the underlying audio processing functions.
For the spectrometer you've got
the issue of multiple sinks for
data, but it is not a problem.
Introduce a dummy link to it from
the real sink. The dummy link can
cause a request for data that is not
honoured. As long as the dummy link knows
it is a dummy and does not care about
the lack of data, everything will be
OK. This is a standard technique for reducing multiple sinks or sources to a single one.
With this kind of network you do not want to do the same calculation twice in one complete update. For example if you mix a high-passed and low-passed version of a signal you do not want to evaluate the original signal twice. You must do something like record a timer tick value with each buffer, and stop propagation of pulls when you see the current tick value is already present. This same mechanism will also protect you against feedback loops in evaluation.
So, those two issues of concern to you are easily addressed within your current framework.
Rate matching where there are different packet rates in different parts of the network is where the problems with the current approach will start. If you are writing audio to disk then for efficiency you'll want to write large chunks infrequently. You don't want to be blocking your servicing of the more frequent small audio input and output processing packets during those writes. A single rate pulling or pushing strategy on its own won't be enough.
Just accept that at some point you may need a more sophisticated way of updating than a single rate network. When that happens you'll need threads for the different rates that are running, or you'll write your own simple scheduler, possibly as simple as calling less frequently evaluated functions one time in n, to make the rates match. You don't need to plan ahead for this. Your audio functions are almost certainly already delegating responsibility for ensuring their input buffers are ready to other functions, and it will only be those other functions that need to change, not the audio functions themselves.
The one thing I would advise at this stage is to be careful to centralise audio buffer
allocation, noticing that buffers are like fenceposts. They don't belong to an audio
function, they lie between the audio functions. Centralising the buffer allocation will make it easy to retrospectively modify the update strategy for different rates in different parts of the network.