I want to use Google translate's v3 PHP library to translate some text into a bunch of different languages. There may be workarounds (though none ideal that I know of), but I'm also just trying to learn.
I wanted to use using multiple calls to translateText, one call per target language. However, to make things faster, I would need to do these requests concurrently, so I was looking into some concurrency options. I was wanting to use calls to translateText instead of constructing a bunch of curl requests manually using curl multi.
I tried the first code example I found from one of the big concurrency libraries I've seen recommended, amphp. I used the function parallelMap, but I'm getting timeout errors when creating processes. I'd guess that I'm probably forking out too many processes at a time.
I'd love to learn if there is an easy way to do concurrency in PHP without having to make a bunch of decisions about how many processes to have running at a time, whether I should use threads vs processes, profiling memory usage, and what PHP thread extension is even any good / if the one I've heard of called "parallel" may be discontinued (as suggested in a comment here).
Any stack overflow post I've found so far is just a link to one giant concurrency library or another that I don't want to have to read a bunch of documentation for. I'd be interested to hear how concurrency like this is normally done / what options there are. I've found many people claim that processes aren't much slower than threads these days, and I can't find quick Google answers as to whether they take a lot more memory than threads. I'm not even positive that a lack of memory is my problem, but it probably is. There has been more complexity involved than I would have expected.
I'm wondering how this is normally handled.
Do people normally just use a pool of workers (processes by default, or threads using, say the "parallel" PHP extension) and set a max number of processes to run at a time to make concurrent requests?
If in a hurry, do people just kind of pick a number that isn't very optimized for how many worker processes to use?
It would be nice if the number of workers to use was set dynamically for you based on how much RAM was available or something, but I guess that's not realistic, since the amount of RAM available can quickly change.
So, would you just need to set up a profiler to see how much ram one worker process/thread uses, or otherwise just make some sort of educated guess as to how many worker processes/threads to use?
Some background:
I am building a server application in php that will need to execute a number of independent tasks on a user request. Theres is a severe requirement on speed for my application so I would like to execute all of those tasks in parallel.
I've looked at several solutions (e.g gearman, rabbitMQ, zeroMQ) and I've decided to go with zeroMQ (fast, good docs, flexible, and doesn't require a broker). This solves the communication/sync problem between the threads for me.
Question:
I would like to initiate the tasks only when the server receives a request (not to have a long running process). So I receive a request -> start parallel computation -> return the result of the computation to the client. One solution for that seems to be pcntl_fork however the docs mention that there ares some issues with using it in a production server env but doesn't really specify what they are?
My other option is to use proc_open, but I like it less because it would require me to serialize the inputs in some way which seems less flexible and fast then forking. Does it have any advantages over pcntl_fork?
Is there another solution (still using php :p)?
Tread carefully, I see several red-flags in your question that lead me to believe you are concerned about things that maybe you don't need to be, and you probably aren't concerned with things you should be.
You say you have a severe requirement for speed - have you validated that normal single threaded PHP is not fast enough? Run any benchmarks, figured out your bottlenecks? If your speed requirement is that great, you might even consider using a different language, for all of PHP's charms it's never going to be the most efficient hammer in the toolbox. Java is a good option for all-out speed, and node.js is a good option if your bottlenecks are IO dependent. My main concern is that, absent more information, this question smells of premature optimization. This may be unfair and you may have omitted those details because it wasn't the heart of your question, but as an outsider I at least wanted to make sure that you think about these things if you haven't already.
You want to avoid long-running processes - why? There's nothing inherently wrong with long-running processes - but it does feel wrong when what you're used to is the pseudo-efficient "on-demand" nature of Apache+mod_php. Be sure you're not trying to avoid something just because you're not used to it.
What you seem to be describing is performing parallel processing from within your PHP web-app - just like any other web page you write, Apache initiates your PHP script, that script forks another process and, rather than performing its actions serially, performs them in parallel, completes and returns to the user at the completion of the page-render. If that's correct, then here is the answer to your original question:
You cannot use pcntl_fork from within a web process, only from the command line. The details of this are on the page you linked to, down in the comments:
It's not a matter of "should not", it's "can not". Even though I have compiled in PCNTL with --enable-pcntl, it turns out that it only compiles in to the CLI version of PHP, not the Apache module. [...] function_exists('pcntl_fork') was returning false even though it compiled correctly. It turns out it returns true just fine from the CLI, and only returns false for HTTP requests. The same is true of ALL of the pcntl_*() functions.
... which means that either you'll have to initiate your forking process as a separate long-running process, or you'll have to start it on demand with proc_open, there is no way to get it to work the way I assume you want it to.
PHP apparently supports now a Thread class. There is also a Mutex and Cond class. One need nevertheless to compile it with the good options and add a PECL extension to have it.
Is there a good tutorial where to start ? Something that shows clearly the fair use of each Class / Method.
I just don't know how to deal with these classes. I would need some concrete examples of doing very simple things that exhibits as many methods as possible.
Thank you.
It has been released as is available on pecl, windows downloads on github pages ... git (source) if you can, more recent code ...
http://pthreads.org contains some more information, and checkout the examples folder too ...
If you're more specific in what you need to do I can provide a precise answer, feel free to contact me directly ...
Sorry about the lack of documentation, I am aggressively developing pthreads in what little time I have around work - it was written while I was on holiday from work and I am now back at work - the documentation will be updated as much as possible as soon as I find the time.
Threading is usually a complicated thing, but it needn't be ... ever seen a thread pool that is less than 100 lines ?? https://github.com/krakjoe/pthreads/blob/master/examples/Pooling.php ... there's one !!
You needn't know about conditions or mutex to operate threads in PHP, pthreads includes easy peasy everything ...
Jump on board, without question, I will perfect it ...
Threading is not easy, in any language. It takes a different train of thought (non-linear). Since what you mentioned is not available in any released PHP version, I would recommend playing with the pctnl and shmop functionality. Specifically pcntl_fork. Both have been around a while and are supported. It is multi-processing and not threading, but may accomplish what you are looking for. But threading is more efficient.
Typically you would use forking to have a parent process check for work, delegate it to a forked process, and then check for more work while the child process does the work. The parent process can check on the status of the child processes. For example, you could make sure you only fork a certain number of child processes and wait for one to complete before forking another.
I've forked up to 500 processes at which point the CPU load got to high. The parent process would check system load and not fork any more if the load was too high.
If you really need to, you could use shared memory (shmop functions) with forking to get much of the functionality you would get with threading.
Right now I'm running 50 PHP (in CLI mode) individual workers (processes) per machine that are waiting to receive their workload (job). For example, the job of resizing an image. In workload they receive the image (binary data) and the desired size. The worker does it's work and returns the resized image back. Then it waits for more jobs (it loops in a smart way). I'm presuming that I have the same executable, libraries and classes loaded and instantiated 50 times. Am I correct? Because this does not sound very effective.
What I'd like to have now is one process that handles all this work and being able to use all available CPU cores while having everything loaded only once (to be more efficient). I presume a new thread would be started for each job and after it finishes, the thread would stop. More jobs would be accepted if there are less than 50 threads doing the work. If all 50 threads are busy, no additional jobs are accepted.
I am using a lot of libraries (for Memcached, Redis, MogileFS, ...) to have access to all the various components that the system uses and Python is pretty much the only language apart from PHP that has support for all of them.
Can Python do what I want and will it be faster and more efficient that the current PHP solution?
Most probably - yes. But don't assume you have to do multithreading. Have a look at the multiprocessing module. It already has an implementation of a Pool included, which is what you could use. And it basically solves the GIL problem (multithreading can run only 1 "standard python code" at any time - that's a very simplified explanation).
It will still fork a process per job, but in a different way than starting it all over again. All the initialisations done- and libraries loaded before entering the worker process will be inherited in a copy-on-write way. You won't do more initialisations than necessary and you will not waste memory for the same libarary/class if you didn't actually make it different from the pre-pool state.
So yes - looking only at this part, python will be wasting less resources and will use a "nicer" worker-pool model. Whether it will really be faster / less CPU-abusing, is hard to tell without testing, or at least looking at the code. Try it yourself.
Added: If you're worried about memory usage, python may also help you a bit, since it has a "proper" garbage collector, while in php GC is a not a priority and not that good (and for a good reason too).
Linux has shared libraries, so those 50 php processes use mostly the same libraries.
You don't sound like you even have a problem at all.
"this does not sound very effective." is not a problem description, if anything those words are a problem on their own. Writing code needs a real reason, else you're just wasting time and/or money.
Python is a fine language and won't perform worse than php. Python's multiprocessing module will probably help a lot too. But there isn't much to gain if the php implementation is not completly insane. So why even bother spending time on it when everything works? That is usually the goal, not a reason to rewrite ...
If you are on a sane operating system then shared libraries should only be loaded once and shared among all processes using them. Memory for data structures and connection handles will obviously be duplicated, but the overhead of stopping and starting the systems may be greater than keeping things up while idle. If you are using something like gearman it might make sense to let several workers stay up even if idle and then have a persistent monitoring process that will start new workers if all the current workers are busy up until a threshold such as the number of available CPUs. That process could then kill workers in a LIFO manner after they have been idle for some period of time.
I found this PECL package called threads, but there is not a release yet. And nothing is coming up on the PHP website.
From the PHP manual for the pthreads extension:
pthreads is an Object Orientated API that allows user-land multi-threading in PHP. It includes all the tools you need to create multi-threaded applications targeted at the Web or the Console. PHP applications can create, read, write, execute and synchronize with Threads, Workers and Stackables.
As unbelievable as this sounds, it's entirely true. Today, PHP can multi-thread for those wishing to try it.
The first release of PHP4, 22 May 2000, PHP was shipped with a thread safe architecture - a way for it to execute multiple instances of it's interpreter in separate threads in multi-threaded SAPI ( Server API ) environments. Over the last 13 years, the design of this architecture has been maintained and advanced: It has been in production use on the worlds largest websites ever since.
Threading in user land was never a concern for the PHP team, and it remains as such today. You should understand that in the world where PHP does it's business, there's already a defined method of scaling - add hardware. Over the many years PHP has existed, hardware has got cheaper and cheaper and so this became less and less of a concern for the PHP team. While it was getting cheaper, it also got much more powerful; today, our mobile phones and tablets have dual and quad core architectures and plenty of RAM to go with it, our desktops and servers commonly have 8 or 16 cores, 16 and 32 gigabytes of RAM, though we may not always be able to have two within budget and having two desktops is rarely useful for most of us.
Additionally, PHP was written for the non-programmer, it is many hobbyists native tongue. The reason PHP is so easily adopted is because it is an easy language to learn and write. The reason PHP is so reliable today is because of the vast amount of work that goes into it's design, and every single decision made by the PHP group. It's reliability and sheer greatness keep it in the spot light, after all these years; where it's rivals have fallen to time or pressure.
Multi-threaded programming is not easy for most, even with the most coherent and reliable API, there are different things to think about, and many misconceptions. The PHP group do not wish for user land multi-threading to be a core feature, it has never been given serious attention - and rightly so. PHP should not be complex, for everyone.
All things considered, there are still benefits to be had from allowing PHP to utilize it's production ready and tested features to allow a means of making the most out of what we have, when adding more isn't always an option, and for a lot of tasks is never really needed.
pthreads achieves, for those wishing to explore it, an API that does allow a user to multi-thread PHP applications. It's API is very much a work in progress, and designated a beta level of stability and completeness.
It is common knowledge that some of the libraries PHP uses are not thread safe, it should be clear to the programmer that pthreads cannot change this, and does not attempt to try. However, any library that is thread safe is useable, as in any other thread safe setup of the interpreter.
pthreads utilizes Posix Threads ( even in Windows ), what the programmer creates are real threads of execution, but for those threads to be useful, they must be aware of PHP - able to execute user code, share variables and allow a useful means of communication ( synchronization ). So every thread is created with an instance of the interpreter, but by design, it's interpreter is isolated from all other instances of the interpreter - just like multi-threaded Server API environments. pthreads attempts to bridge the gap in a sane and safe way. Many of the concerns of the programmer of threads in C just aren't there for the programmer of pthreads, by design, pthreads is copy on read and copy on write ( RAM is cheap ), so no two instances ever manipulate the same physical data, but they can both affect data in another thread. The fact that PHP may use thread unsafe features in it's core programming is entirely irrelevant, user threads, and it's operations are completely safe.
Why copy on read and copy on write:
public function run() {
...
(1) $this->data = $data;
...
(2) $this->other = someOperation($this->data);
...
}
(3) echo preg_match($pattern, $replace, $thread->data);
(1) While a read, and write lock are held on the pthreads object data store, data is copied from its original location in memory to the object store. pthreads does not adjust the refcount of the variable, Zend is able to free the original data if there are no further references to it.
(2) The argument to someOperation references the object store, the original data stored, which it itself a copy of the result of (1), is copied again for the engine into a zval container, while this occurs a read lock is held on the object store, the lock is released and the engine can execute the function. When the zval is created, it has a refcount of 0, enabling the engine to free the copy on completion of the operation, because no other references to it exist.
(3) The last argument to preg_match references the data store, a read lock is obtained, the data set in (1) is copied to a zval, again with a refcount of 0. The lock is released, The call to preg_match operates on a copy of data, that is itself a copy of the original data.
Things to know:
The object store's hash table where data is stored, thread safe, is
based on the TsHashTable shipped with PHP, by Zend.
The object store has a read and write lock, an additional access lock is provided for the TsHashTable such that if requires ( and it does, var_dump/print_r, direct access to properties as the PHP engine wants to reference them ) pthreads can manipulate the TsHashTable outside of the defined API.
The locks are only held while the copying operations occur, when the copies have been made the locks are released, in a sensible order.
This means:
When a write occurs, not only are a read and write lock held, but an
additional access lock. The table itself is locked down, there is no
possible way another context can lock, read, write or affect it.
When a read occurs, not only is the read lock held, but the
additional access lock too, again the table is locked down.
No two contexts can physically nor concurrently access the same data from the object store, but writes made in any context with a reference will affect the data read in any context with a reference.
This is shared nothing architecture and the only way to exist is co-exist. Those a bit savvy will see that, there's a lot of copying going on here, and they will wonder if that is a good thing. Quite a lot of copying goes on within a dynamic runtime, that's the dynamics of a dynamic language. pthreads is implemented at the level of the object, because good control can be gained over one object, but methods - the code the programmer executes - have another context, free of locking and copies - the local method scope. The object scope in the case of a pthreads object should be treated as a way to share data among contexts, that is it's purpose. With this in mind you can adopt techniques to avoid locking the object store unless it's necessary, such as passing local scope variables to other methods in a threaded object rather than having them copy from the object store upon execution.
Most of the libraries and extensions available for PHP are thin wrappers around 3rd parties, PHP core functionality to a degree is the same thing. pthreads is not a thin wrapper around Posix Threads; it is a threading API based on Posix Threads. There is no point in implementing Threads in PHP that it's users do not understand or cannot use. There's no reason that a person with no knowledge of what a mutex is or does should not be able to take advantage of all that they have, both in terms of skill, and resources. An object functions like an object, but wherever two contexts would otherwise collide, pthreads provides stability and safety.
Anyone who has worked in java will see the similarities between a pthreads object and threading in java, those same people will have no doubt seen an error called ConcurrentModificationException - as it sounds an error raised by the java runtime if two threads write the same physical data concurrently. I understand why it exists, but it baffles me that with resources as cheap as they are, coupled with the fact the runtime is able to detect the concurrency at the exact and only time that safety could be achieved for the user, that it chooses to throw a possibly fatal error at runtime rather than manage the execution and access to the data.
No such stupid errors will be emitted by pthreads, the API is written to make threading as stable, and compatible as is possible, I believe.
Multi-threading isn't like using a new database, close attention should be paid to every word in the manual and examples shipped with pthreads.
Lastly, from the PHP manual:
pthreads was, and is, an experiment with pretty good results. Any of its limitations or features may change at any time; that is the nature of experimentation. It's limitations - often imposed by the implementation - exist for good reason; the aim of pthreads is to provide a useable solution to multi-tasking in PHP at any level. In the environment which pthreads executes, some restrictions and limitations are necessary in order to provide a stable environment.
Here is an example of what Wilco suggested:
$cmd = 'nohup nice -n 10 /usr/bin/php -c /path/to/php.ini -f /path/to/php/file.php action=generate var1_id=23 var2_id=35 gen_id=535 > /path/to/log/file.log & echo $!';
$pid = shell_exec($cmd);
Basically this executes the PHP script at the command line, but immediately returns the PID and then runs in the background. (The echo $! ensures nothing else is returned other than the PID.) This allows your PHP script to continue or quit if you want. When I have used this, I have redirected the user to another page, where every 5 to 60 seconds an AJAX call is made to check if the report is still running. (I have a table to store the gen_id and the user it's related to.) The check script runs the following:
exec('ps ' . $pid , $processState);
if (count($processState) < 2) {
// less than 2 rows in the ps, therefore report is complete
}
There is a short post on this technique here: http://nsaunders.wordpress.com/2007/01/12/running-a-background-process-in-php/
There is nothing available that I'm aware of. The next best thing would be to simply have one script execute another via CLI, but that's a bit rudimentary. Depending on what you are trying to do and how complex it is, this may or may not be an option.
In short: yes, there is multithreading in php but you should use multiprocessing instead.
Backgroud info: threads vs. processes
There is always a bit confusion about the distinction of threads and processes, so i'll shortly describe both:
A thread is a sequence of commands that the CPU will process. The only data it consists of is a program counter. Each CPU core will only process one thread at a time but can switch between the execution of different ones via scheduling.
A process is a set of shared resources. That means it consists of a part of memory, variables, object instances, file handles, mutexes, database connections and so on. Each process also contains one or more threads. All threads of the same process share its resources, so you may use a variable in one thread that you created in another. If those threads are parts of two different processes, then they cannot access each others resources directly. In this case you need inter-process communication through e.g. pipes, files, sockets...
Multiprocessing
You can achieve parallel computing by creating new processes (that also contain a new thread) with php. If your threads do not need much communication or synchronization, this is your choice, since the processes are isolated and cannot interfere with each other's work. Even if one crashes, that doesn't concern the others. If you do need much communication, you should read on at "multithreading" or - sadly - consider using another programming language, because inter-process communication and synchronization introduces a lot of complexion.
In php you have two ways to create a new process:
let the OS do it for you: you can tell your operation system to create a new process and run a new (or the same) php script in it.
for linux you can use the following or consider Darryl Hein's answer:
$cmd = 'nice php script.php 2>&1 & echo $!';
pclose(popen($cmd, 'r'));
for windows you may use this:
$cmd = 'start "processname" /MIN /belownormal cmd /c "script.php 2>&1"';
pclose(popen($cmd, 'r'));
do it yourself with a fork: php also provides the possibility to use forking through the function pcntl_fork(). A good tutorial on how to do this can be found here but i strongly recommend not to use it, since fork is a crime against humanity and especially against oop.
Multithreading
With multithreading all your threads share their resources so you can easily communicate between and synchronize them without a lot of overhead. On the other side you have to know what you are doing, since race conditions and deadlocks are easy to produce but very difficult to debug.
Standard php does not provide any multithreading but there is an (experimental) extension that actually does - pthreads. Its api documentation even made it into php.net.
With it you can do some stuff as you can in real programming languages :-) like this:
class MyThread extends Thread {
public function run(){
//do something time consuming
}
}
$t = new MyThread();
if($t->start()){
while($t->isRunning()){
echo ".";
usleep(100);
}
$t->join();
}
For linux there is an installation guide right here at stackoverflow's.
For windows there is one now:
First you need the thread-safe version of php.
You need the pre-compiled versions of both pthreads and its php extension. They can be downloaded here. Make sure that you download the version that is compatible with your php version.
Copy php_pthreads.dll (from the zip you just downloaded) into your php extension folder ([phpDirectory]/ext).
Copy pthreadVC2.dll into [phpDirectory] (the root folder - not the extension folder).
Edit [phpDirectory]/php.ini and insert the following line
extension=php_pthreads.dll
Test it with the script above with some sleep or something right there where the comment is.
And now the big BUT: Although this really works, php wasn't originally made for multithreading. There exists a thread-safe version of php and as of v5.4 it seems to be nearly bug-free but using php in a multi-threaded environment is still discouraged in the php manual (but maybe they just did not update their manual on this, yet). A much bigger problem might be that a lot of common extensions are not thread-safe. So you might get threads with this php extension but the functions you're depending on are still not thread-safe so you will probably encounter race conditions, deadlocks and so on in code you did not write yourself...
You can use pcntl_fork() to achieve something similar to threads. Technically it's separate processes, so the communication between the two is not as simple with threads, and I believe it will not work if PHP is called by apache.
If anyone cares, I have revived php_threading (not the same as threads, but similar) and I actually have it to the point where it works (somewhat) well!
Project page
Download (for Windows PHP 5.3 VC9 TS)
Examples
README
pcntl_fork() is what you are searching for, but its process forking not threading.
so you will have the problem of data exchange. to solve them you can use phps semaphore functions ( http://www.php.net/manual/de/ref.sem.php ) message queues may be a bit easier for the beginning than shared memory segments.
Anyways, a strategy i am using in a web framework that i am developing which loads resource intensive blocks of a web page (probably with external requests) parallel:
i am doing a job queue to know what data i am waiting for and then i fork off the jobs for every process. once done they store their data in the apc cache under a unique key the parent process can access. once every data is there it continues.
i am using simple usleep() to wait because inter process communication is not possible in apache (children will loose the connection to their parents and become zombies...).
so this brings me to the last thing:
its important to self kill every child!
there are as well classes that fork processes but keep data, i didn't examine them but zend framework has one, and they usually do slow but reliably code.
you can find it here:
http://zendframework.com/manual/1.9/en/zendx.console.process.unix.overview.html
i think they use shm segments!
well last but not least there is an error on this zend website, minor mistake in the example.
while ($process1->isRunning() && $process2->isRunning()) {
sleep(1);
}
should of course be:
while ($process1->isRunning() || $process2->isRunning()) {
sleep(1);
}
There is a Threading extension being activley developed based on PThreads that looks very promising at https://github.com/krakjoe/pthreads
Just an update, its seem that PHP guys are working on supporting thread and its available now.
Here is the link to it:
http://php.net/manual/en/book.pthreads.php
I have a PHP threading class that's been running flawlessly in a production environment for over two years now.
EDIT: This is now available as a composer library and as part of my MVC framework, Hazaar MVC.
See: https://git.hazaarlabs.com/hazaar/hazaar-thread
I know this is a way old question, but you could look at http://phpthreadlib.sourceforge.net/
Bi-directional communication, support for Win32, and no extensions required.
Ever heard about appserver from techdivision?
It is written in php and works as a appserver managing multithreads for high traffic php applications. Is still in beta but very promesing.
There is the rather obscure, and soon to be deprecated, feature called ticks. The only thing I have ever used it for, is to allow a script to capture SIGKILL (Ctrl+C) and close down gracefully.