I have a long running (5-10 hours) Mac app that processes 5000 items. Each item is processed by performing a number of transforms (using Saxon), running a bunch of scripts (in Python and Racket), collecting data, and serializing it as a set of XML files, a SQLite database, and a CoreData database. Each item is completely independent from every other item.
In summary, it does a lot, takes a long time, and appears to be highly parallelizable.
After loading up all the items that need processing it, the app uses GCD to parallelize the work, using dispatch_apply:
dispatch_apply(numberOfItems, dispatch_get_global_queue(DISPATCH_QUEUE_PRIORITY_HIGH, 0), ^(size_t i) {
#autoreleasepool {
...
}
});
I'm running the app on a Mac Pro with 12 cores (24 virtual). So I would expect to have 24 items being processed at all times. However, I found through logging that the number of items being processed varies between 8 and 24. This is literally adding hours to the run time (assuming it could work on 24 items at a time).
On the one hand, perhaps GCD is really, really smart and it is already giving me the maximum throughput. But I'm worried that, because much of the work happens in scripts that are spawned by this app, maybe GCD is reasoning from incomplete information and isn't making the best decisions.
Any ideas how to improve performance? After correctness, the number one desired attribute is shortening how long it takes this app to run. I don't care about power consumption, hogging the Mac Pro, or anything else.
UPDATE: In fact, this looks alarming in the docs: "The actual number of tasks executed by a concurrent queue at any given moment is variable and can change dynamically as conditions in your application change. Many factors affect the number of tasks executed by the concurrent queues, including the number of available cores, the amount of work being done by other processes, and the number and priority of tasks in other serial dispatch queues." (emphasis added) It looks like having other processes doing work will adversely affect scheduling in the app.
It'd be nice to be able to just say "run these blocks concurrently, one per core, don't try to do anything smarter".
If you are bound and determined, you can explicitly spawn 24 threads using the NSThread API, and have each of those threads pull from a synchronized queue of work items. I would bet money that performance would get noticeably worse.
GCD works at its most efficient when the work items submitted to it never block. That said, the workload you're describing is rather complex and rife with opportunities for your threads to block. For starters, you're spawning a bunch of other processes. Right here, this means that you're already relying on the OS to divvy up time/resources between your master task and these slave tasks. Other than setting the OS priority of each subprocess, the OS scheduler has no way to know which processes are more important than others, and by default, your subprocesses are going to have the same priority as their parent. That said, it doesn't sound like you have anything to gain by tweaking process priorities. I'm assuming you're blocking the master task thread that's waiting for the slave tasks to complete. That is effectively parking that thread -- it can do no useful work. But like I said, I don't think there's much to be gained by tweaking the OS priorities of your slave tasks, because this really sounds like it's an I/O bound workflow...
You go on to describe three I/O-heavy operations ("serializing it as a set of XML files, a SQLite database, and a CoreData database.") So now you have all these different threads and processes vying for what is presumably a shared bulk storage device. (i.e. unless you're writing to 24 different databases, on 24 separate hard drives, one for each core, your process is ultimately going to be serialized at the disk accesses.) Even if you had 24 different hard drives, writing to a hard drive (even an SSD) is comparatively slow. Your threads are going to be taken off of the CPU they were running on (so that another thread that's waiting can run) for virtually any blocking disk write.
If you wanted to maximize the performance you're getting out of GCD, you would probably want to rewrite all the stuff you're doing in subtasks in C/C++/Objective-C, bringing them in-process, and then conducting all the associated I/O using dispatch_io primitives. For API where you don't control the low-level reads and writes, you would want to carefully manage and tune your workload to optimize it for the hardware you have. For instance, if you have a bunch of stuff to write to a single, shared SQLite database, there's no point in ever having more than one thread trying to write to that database at once. You'd be better off making one thread (or a serial GCD queue) to write to SQLite and submitting tasks to that after pre-processing is done.
I could go on for quite a while here, but the bottom line is that you've got a complex, seemingly I/O bound workflow here. At the highest-level, CPU utilization or "number of running threads" is going to be a particularly poor measure of performance for such a task. By using sub-processes (i.e. scripts), you're putting a lot of control into the hands of the OS, which knows effectively nothing about your workload a priori, and therefore can do nothing except use its general scheduler to divvy up resources. GCD's opaque thread pool management is really the least of your problems.
On a practical level, if you want to speed things up, go buy multiple, faster (i.e. SSD) hard drives, and rework your task/workflow to utilize them separately and in parallel. I suspect that would yield the biggest bang for your buck (for some equivalence relation of time == money == hardware.)
Related
I have a question about how scheduling is done. I know that when a system has multiple CPUs scheduling is usually done on a per processor bases. Each processor runs its own scheduler accessing a ready list of only those processes that are running on it.
So what would be the pros and cons when compared to an approach where there is a single ready list that all processors share?
Like what issues are there when assigning processes to processors and what issues might be caused if a process always lives on one processor? In terms of the mutex locking of data structures and time spent waiting on for the locks are there any issues to that?
Generally there is one, giant problem when it comes to multi-core CPU systems - cache coherency.
What does cache coherency mean?
Access to main memory is hard. Depending on the memory frequency, it can take between a few thousand to a few million cycles to access some data in RAM - that's a whole lot of time the CPU is doing no useful work. It'd be significantly better if we minimized this time as much as possible, but the hardware required to do this is expensive, and typically must be in very close proximity to the CPU itself (we're talking within a few millimeters of the core).
This is where the cache comes in. The cache keeps a small subset of main memory in close proximity to the core, allowing accesses to this memory to be several orders of magnitude faster than main memory. For reading this is a simple process - if the memory is in the cache, read from cache, otherwise read from main memory.
Writing is a bit more tricky. Writing to the cache is fast, but now main memory still holds the original value. We can update that memory, but that takes a while, sometimes even longer than reading depending on the memory type and board layout. How do we minimize this as well?
The most common way to do so is with a write-back cache, which, when written to, will flush the data contained in the cache back to main memory at some later point when the CPU is idle or otherwise not doing something. Depending on the CPU architecture, this could be done during idle conditions, or interleaved with CPU instructions, or on a timer (this is up to the designer/fabricator of the CPU).
Why is this a problem?
In a single core system, there is only one path for reads and writes to take - they must go through the cache on their way to main memory, meaning the programs running on the CPU only see what they expect - if they read a value, modified it, then read it back, it would be changed.
In a multi-core system, however, there are multiple paths for data to take when going back to main memory, depending on the CPU that issued the read or write. this presents a problem with write-back caching, since that "later time" introduces a gap in which one CPU might read memory that hasn't yet been updated.
Imagine a dual core system. A job starts on CPU 0 and reads a memory block. Since the memory block isn't in CPU 0's cache, it's read from main memory. Later, the job writes to that memory. Since the cache is write-back, that write will be made to CPU 0's cache and flushed back to main memory later. If CPU 1 then attempts to read that same memory, CPU 1 will attempt to read from main memory again, since it isn't in the cache of CPU 1. But the modification from CPU 0 hasn't left CPU 0's cache yet, so the data you get back is not valid - your modification hasn't gone through yet. Your program could now break in subtle, unpredictable, and potentially devastating ways.
Because of this, cache synchronization is done to alleviate this. Application IDs, address monitoring, and other hardware mechanisms exist to synchronize the caches between multiple CPUs. All of these methods have one common problem - they all force the CPU to take time doing bookkeeping rather than actual, useful computations.
The best method of avoiding this is actually keeping processes on one processor as much as possible. If the process doesn't migrate between CPUs, you don't need to keep the caches synchronized, as the other CPUs won't be accessing that memory at the same time (unless the memory is shared between multiple processes, but we'll not go into that here).
Now we come to the issue of how to design our scheduler, and the three main problems there - avoiding process migration, maximizing CPU utilization, and scalability.
Single Queue Multiprocessor scheduling (SQMS)
Single Queue Multiprocessor schedulers are what you suggested - one queue containing available processes, and each core accesses the queue to get the next job to run. This is fairly simple to implement, but has a couple of major drawbacks - it can cause a whole lot of process migration, and does not scale well to larger systems with more cores.
Imagine a system with four cores and five jobs, each of which takes about the same amount of time to run, and each of which is rescheduled when completed. On the first run through, CPU 0 takes job A, CPU 1 takes B, CPU 2 takes C, and CPU 3 takes D, while E is left on the queue. Let's then say CPU 0 finishes job A, puts it on the back of the shared queue, and looks for another job to do. E is currently at the front of the queue, to CPU 0 takes E, and goes on. Now, CPU 1 finishes job B, puts B on the back of the queue, and looks for the next job. It now sees A, and starts running A. But since A was on CPU 0 before, CPU 1 now needs to sync its cache with CPU 0, resulting in lost time for both CPU 0 and CPU 1. In addition, if two CPUs both finish their operations at the same time, they both need to write to the shared list, which has to be done sequentially or the list will get corrupted (just like in multi-threading). This requires that one of the two CPUs wait for the other to finish their writes, and sync their cache back to main memory, since the list is in shared memory! This problem gets worse and worse the more CPUs you add, resulting in major problems with large servers (where there can be 16 or even 32 CPU cores), and being completely unusable on supercomputers (some of which have upwards of 1000 cores).
Multi-queue Multiprocessor Scheduling (MQMS)
Multi-queue multiprocessor schedulers have a single queue per CPU core, ensuring that all local core scheduling can be done without having to take a shared lock or synchronize the cache. This allows for systems with hundreds of cores to operate without interfering with one another at every scheduling interval, which can happen hundreds of times a second.
The main issue with MQMS comes from CPU Utilization, where one or more CPU cores is doing the majority of the work, and scheduling fairness, where one of the processes on the computer is being scheduled more often than any other process with the same priority.
CPU Utilization is the biggest issue - no CPU should ever be idle if a job is scheduled. However, if all CPUs are busy, so we schedule a job to a random CPU, and a different CPU ends up becoming idle, it should "steal" the scheduled job from the original CPU to ensure every CPU is doing real work. Doing so, however, requires that we lock both CPU cores and potentially sync the cache, which may degrade any speedup we could get by stealing the scheduled job.
In conclusion
Both methods exist in the wild - Linux actually has three different mainstream scheduler algorithms, one of which is an SQMS. The choice of scheduler really depends on the way the scheduler is implemented, the hardware you plan to run it on, and the types of jobs you intend to run. If you know you only have two or four cores to run jobs, SQMS is likely perfectly adequate. If you're running a supercomputer where overhead is a major concern, then an MQMS might be the way to go. For a desktop user - just trust the distro, whether that's a Linux OS, Mac, or Windows. Generally, the programmers for the operating system you've got have done their homework on exactly what scheduler will be the best option for the typical use case of their system.
This whitepaper describes the differences between the two types of scheduling algorithms in place.
Vulkan is intended to be thin and explicit to user, but queues are a big exception to this rule: queues may be multiplexed by driver and it's not always obvious if using multiple queues from a family will improve performance or not.
After one of driver updates, I've got 2 transfer-only queues instead of one, but I'm pretty sure that there will be no benefit in using them in parallel for data streaming compared to just using one of them (will be happy to be proved wrong)
So why not just say "we have N separate hardware queues and if you want to use some of them in parallel, just mutex it yourself"? Now it looks like there's no way to know, how independent queues in family really are.
GPUs these days have to contend with a multi-processed world. Different programs can access the same hardware, and GPUs have to be able to deal with that. As such, having parallel input streams for a single piece of actual hardware is no different from being able to create more CPU threads than you have actual CPU cores.
That is, a queue from a family is probably not "mutexing" access to the actual hardware. At least, not in a CPU way. If multiple queues from a family are different paths to execute stuff on the same hardware, then the way that hardware gets populated from these multiple queues probably happens at the GPU level. That is, it's an actual hardware feature.
And you could never get performance equivalent to that hardware feature by "mutexing it yourself". For example:
I've got 2 transfer-only queues instead of one, but I'm pretty sure that there will be no benefit in using them in parallel for data streaming compared to just using one of them
Let's assume that there really is only one hardware DMA channel with a fixed bandwidth behind that transfer queue. This means that, at any one time, only one thing can be DMA'd from CPU memory to GPU memory at one time.
Now, let's say you have some DMA work to do. You want to upload a bunch of stuff. But every now and then, you need to download some rendering product. And that download needs to complete ASAP, because you need to reuse the image that stores those bytes.
With prioritized queues, you can give the download transfer queue much higher priority than the upload queue. If the hardware permits it, then it can interrupt the upload to perform the download, then get back to the upload.
With your way, you'd have to upload each item one at a time at regular intervals. A process that will have to be able to be interrupted by a possible download. To do that, you'd basically have to have a recurring tasks that shows up to perform and submit a single upload to the transfer queue.
It'd be much more efficient to just throw the work at the GPU and let its priority system take care of it. Even if there is no priority system, then it'll probably perform operations round-robin, jumping back and forth between the input transfer queue operations rather than waiting for one queue to run dry before trying another.
But of course, this is all hypothetical. You'd need to do profiling work to make sure that these things pan out.
The main issue with queues within families is that they sometimes represent distinct hardware with their own dedicated resources and sometimes they don't. AMD's hardware for example offers two transfer queues, but these actually use separate DMA channels. Granted, they probably still share the same overall bandwidth, but it's not a simple case of one queue having to wait to execute work until the other queue has executed a transfer command.
I have been confused about the issue of context switches between processes, given round robin scheduler of certain time slice (which is what unix/windows both use in a basic sense).
So, suppose we have 200 processes running on a single core machine. If the scheduler is using even 1ms time slice, each process would get its share every 200ms, which is probably not the case (imagine a Java high-frequency app, I would not assume it gets scheduled every 200ms to serve requests). Having said that, what am I missing in the picture?
Furthermore, java and other languages allows to put the running thread to sleep for e.g. 100ms. Am I correct in saying that this does not cause context switch, and if so, how is this achieved?
So, suppose we have 200 processes running on a single core machine. If
the scheduler is using even 1ms time slice, each process would get its
share every 200ms, which is probably not the case (imagine a Java
high-frequency app, I would not assume it gets scheduled every 200ms
to serve requests). Having said that, what am I missing in the
picture?
No, you aren't missing anything. It's the same case in the case of non-pre-emptive systems. Those having pre-emptive rights(meaning high priority as compared to other processes) can easily swap the less useful process, up to an extent that a high-priority process would run 10 times(say/assume --- actual results are totally depending on the situation and implementation) than the lowest priority process till the former doesn't produce the condition of starvation of the least priority process.
Talking about the processes of similar priority, it totally depends on the Round-Robin Algorithm which you've mentioned, though which process would be picked first is again based on the implementation. And, Windows and Unix have same process scheduling algorithms. Windows and Unix does utilise Round-Robin, but, Linux task scheduler is called Completely Fair Scheduler (CFS).
Furthermore, java and other languages allows to put the running thread
to sleep for e.g. 100ms. Am I correct in saying that this does not
cause context switch, and if so, how is this achieved?
Programming languages and libraries implement "sleep" functionality with the aid of the kernel. Without kernel-level support, they'd have to busy-wait, spinning in a tight loop, until the requested sleep duration elapsed. This would wastefully consume the processor.
Talking about the threads which are caused to sleep(Thread.sleep(long millis)) generally the following is done in most of the systems :
Suspend execution of the process and mark it as not runnable.
Set a timer for the given wait time. Systems provide hardware timers that let the kernel register to receive an interrupt at a given point in the future.
When the timer hits, mark the process as runnable.
I hope you might be aware of threading models like one to one, many to one, and many to many. So, I am not getting into much detail, jut a reference for yourself.
It might appear to you as if it increases the overhead/complexity. But, that's how threads(user-threads created in JVM) are operated upon. And, then the selection is based upon those memory models which I mentioned above. Check this Quora question and answers to that one, and please go through the best answer given by Robert-Love.
For further reading, I'd suggest you to read from Scheduling Algorithms explanation on OSDev.org and Operating System Concepts book by Galvin, Gagne, Silberschatz.
I searched a variety of sources but don't really understand the difference between using NSThreads and GCD. I'm completely new to the OS X platform so I might be completely misinterpreting this.
From what I read online, GCD seems to do the exact same thing as basic threads (POSIX, NSThreads etc.) while adding much more technical jargon ("blocks"). It seems to just overcomplicate the basic thread creation system (create thread, run function).
What exactly is GCD and why would it ever be preferred over traditional threading? When should traditional threads be used rather than GCD? And finally is there a reason for GCD's strange syntax? ("blocks" instead of simply calling functions).
I am on Mac OS X 10.6.8 Snow Leopard and I am not programming for iOS - I am programming for Macs. I am using Xcode 3.6.8 in Cocoa, creating a GUI application.
Advantages of Dispatch
The advantages of dispatch are mostly outlined here:
Migrating Away from Threads
The idea is that you eliminate work on your part, since the paradigm fits MOST code more easily.
It reduces the memory penalty your application pays for storing thread stacks in the application’s memory space.
It eliminates the code needed to create and configure your threads.
It eliminates the code needed to manage and schedule work on threads.
It simplifies the code you have to write.
Empirically, using GCD-type locking instead of #synchronized is about 80% faster or more, though micro-benchmarks may be deceiving. Read more here, though I think the advice to go async with writes does not apply in many cases, and it's slower (but it's asynchronous).
Advantages of Threads
Why would you continue to use Threads? From the same document:
It is important to remember that queues are not a panacea for
replacing threads. The asynchronous programming model offered by
queues is appropriate in situations where latency is not an issue.
Even though queues offer ways to configure the execution priority of
tasks in the queue, higher execution priorities do not guarantee the
execution of tasks at specific times. Therefore, threads are still a
more appropriate choice in cases where you need minimal latency, such
as in audio and video playback.
Another place where I haven't personally found an ideal solution using queues is daemon processes that need to be constantly rescheduled. Not that you cannot reschedule them, but looping within a NSThread method is simpler (I think). Edit: Now I'm convinced that even in this context, GCD-style locking would be faster, and you could also do a loop within a GCD-dispatched operation.
Blocks in Objective-C?
Blocks are really horrible in Objective-C due to the awful syntax (though Xcode can sometimes help with autocompletion, at least). If you look at blocks in Ruby (or any other language, pretty much) you'll see how simple and elegant they are for dispatching operations. I'd say that you'll get used to the Objective-C syntax, but I really think that you'll get used to copying from your examples a lot :)
You might find my examples from here to be helpful, or just distracting. Not sure.
While the answers so far are about the context of threads vs GCD inside the domain of a single application and the differences it has for programming, the reason you should always prefer GCD is because of multitasking environments (since you are on MacOSX and not iOS). Threads are ok if your application is running alone on your machine. Say, you have a video edition program and want to apply some effect to the video. The render is going to take 10 minutes on a machine with eight cores. Fine.
Now, while the video app is churning in the background, you open an image edition program and play with some high resolution image, decide to apply some special image filter and your image application being clever detects you have eight cores and starts eight threads to process the image. Nice isn't it? Except that's terrible for performance. The image edition app doesn't know anything about the video app (and vice versa) and therefore both will request their respectively optimum number of threads. And there will be pain and blood while the cores try to switch from one thread to another, because to avoid starvation the CPU will eventually let all threads run, even though in this situation it would be more optimal to run only 4 threads for the video app and 4 threads for the image app.
For a more detailed reference, take a look at http://deusty.blogspot.com/2010/11/introducing-gcd-based-cocoahttpserver.html where you can see a benchmark of an HTTP server using GCD versus thread, and see how it scales. Once you understand the problem threads have for multicore machines in multi-app environments, you will always want to use GCD, simply because threads are not always optimal, while GCD potentially can be since the OS can scale thread usage per app depending on load.
Please, remember we won't have more GHz in our machines any time soon. From now on we will only have more cores, so it's your duty to use the best tool for this environment, and that is GCD.
Blocks allow for passing a block of code to execute. Once you get past the "strange syntax", they are quite powerful.
GCD also uses queues which if used properly can help with lock free concurrency if the code executing in the separate queues are isolated. It's a simpler way to offer background and concurrency while minimizing the chance for deadlocks (if used right).
The "strange syntax" is because they chose the caret (^) because it was one of the few symbols that wasn't overloaded as an operator in C++
See:
https://developer.apple.com/library/ios/#documentation/General/Conceptual/ConcurrencyProgrammingGuide/OperationQueues/OperationQueues.html
When it comes to adding concurrency to an application, dispatch queues
provide several advantages over threads. The most direct advantage is
the simplicity of the work-queue programming model. With threads, you
have to write code both for the work you want to perform and for the
creation and management of the threads themselves. Dispatch queues let
you focus on the work you actually want to perform without having to
worry about the thread creation and management. Instead, the system
handles all of the thread creation and management for you. The
advantage is that the system is able to manage threads much more
efficiently than any single application ever could. The system can
scale the number of threads dynamically based on the available
resources and current system conditions. In addition, the system is
usually able to start running your task more quickly than you could if
you created the thread yourself.
Although you might think rewriting your code for dispatch queues would
be difficult, it is often easier to write code for dispatch queues
than it is to write code for threads. The key to writing your code is
to design tasks that are self-contained and able to run
asynchronously. (This is actually true for both threads and dispatch
queues.)
...
Although you would be right to point out that two tasks running in a
serial queue do not run concurrently, you have to remember that if two
threads take a lock at the same time, any concurrency offered by the
threads is lost or significantly reduced. More importantly, the
threaded model requires the creation of two threads, which take up
both kernel and user-space memory. Dispatch queues do not pay the same
memory penalty for their threads, and the threads they do use are kept
busy and not blocked.
GCD (Grand Central Dispatch): GCD provides and manages FIFO queues to which your application can submit tasks in the form of block objects. Work submitted to dispatch queues are executed on a pool of threads fully managed by the system. No guarantee is made as to the thread on which a task executes. Why GCD over threads :
How much work your CPU cores are doing
How many CPU cores you have.
How much threads should be spawned.
If GCD needs it can go down into the kernel and communicate about resources, thus better scheduling.
Less load on kernel and better sync with OS
GCD uses existing threads from thread pool instead of creating and then destroying.
Best advantage of the system’s hardware resources, while allowing the operating system to balance the load of all the programs currently running along with considerations like heating and battery life.
I have shared my experience with threads, operating system and GCD AT http://iosdose.com
Is it safe? For instance, if I create a bunch of different GCD queues that each compress (tar cvzf) some files, am I doing something wrong? Will the hard drive be destroyed?
Or does the system properly take care of such things?
Dietrich's answer is correct save for one detail (that is completely non-obvious).
If you were to spin off, say, 100 asynchronous tar executions via GCD, you'd quickly find that you have 100 threads running in your application (which would also be dead slow due to gross abuse of the I/O subsystem).
In a fully asynchronous concurrent system with queues, there is no way to know if a particular unit of work is blocked because it is waiting for a system resource or waiting for some other enqueued unit of work. Therefore, anytime anything blocks, you pretty much have to spin up another thread and consume another unit of work or risk locking up the application.
In such a case, the "obvious" solution is to wait a bit when a unit of work blocks before spinning up another thread to de-queue and process another unit of work with the hope that the first unit of work "unblocks" and continues processing.
Doing so, though, would mean that any asynchronous concurrent system with interaction between units of work -- a common case -- would be so slow as to be useless.
Far more effective is to limit the # of units of work that are enqueued in the global asynchronous queues at any one time. A GCD semaphore makes this quite easy; you have a single serial queue into which all units of work are enqueued. Every time you dequeue a unit of work, you increment the semaphore. Every time a unit of work is completed, you decrement the semaphore. As long as the semaphore is below some maximum value (say, 4), then you enqueue a new unit of work.
If you take something that is normally IO limited, such as tar, and run a bunch of copies in GCD,
It will run more slowly because you are throwing more CPU at an IO-bound task, meaning the IO will be more scattered and there will be more of it at the same time,
No more than N tasks will run at a time, which is the point of GCD, so "a billion queue entries" and "ten queue entries" give you the same thing if you have less than 10 threads,
Your hard drive will be fine.
Even though this question was asked back in May, it's still worth noting that GCD has now provided I/O primitives with the release of 10.7 (OS X Lion). See the man pages for dispatch_read and dispatch_io_create for examples on how to do efficient I/O with the new APIs. They are smart enough to properly schedule I/O against a single disk (or multiple disks) with knowledge of how much concurrency is, or is not, possible in the actual I/O requests.