I'm having problems scaling up an application that uses the actor model and zeromq. To put it simply: I'm trying to create thousands of threads that communicate via sockets. Similar to what one would do with a Erlang-type message passing. I'm not doing it for multicore/performance reasons, but because framing it in this way gives me very clean code.
From a philosophical point of view it sounds as if this is what zmq developers would like to achieve, e.g.
http://zeromq.org/whitepapers:multithreading-magic
However, it seems as if there are some practical limitations. At 1024 inproc sockets I start getting the "ZMQError: Too many open files" error. TCP gives me the typical "Assertion failed: fds.size () <= FD_SETSIZE" crash.
Why does inproc sockets have this limit?
To get it to work I've had to group together items to share a socket. Is there a better way?
Is zmq just the wrong tool for this kind of job? i.e. it's still more a network library than an actor message passing library?
ZMQ uses file descriptors as the "resource unit" for inproc connections. There is a limit for file descriptors set by the OS, you should be able to modify that (found several potential avenues for Windows with a quick Google search), though I don't know what the performance impact might be.
It looks like this is related to the ZMQ library using C code that is portable among systems for opening new files, rather than Windows native code that doesn't suffer from this same limitation.
Related
I am testing .NET version of ZeroMQ to understand how to handle network failures. I put the server (pub socket) to one external machine and debugging the client (sub socket). If I stop my local Wi-Fi connection for seconds, then ZeroMQ automatically recovers and I even get remaining values. However, if I disable Wi-Fi for longer time like a minute, then it just gets stuck on a frame waiting. How can I configure this period when ZeroMQ is still able to recover? And how can I reconnect manually after, say, several minutes? How can I understand that the socket is locked and I need to kill/open again?
Q :" How can I configure this ... ?"
A :Use the .NET versions of zmq_setsockopt() detailed parameter settings - family of link-management parameters alike ZMQ_RECONNECT_IVL, ZMQ_RCVTIMEO and the likes.
All other questions depend on your code.
If using blocking-forms of the .recv()-methods, you can easily throw yourself into unsalvageable deadlocks, best never block your own code ( why one would ever deliberately lose one's own code domain-of-control ).
If in a need to indeed understand low-level internal link-management details, do not hesitate to use zmq_socket_monitor() instrumentation ( if not available in .NET binding, still may use another language to see details the monitor-instance reports about link-state and related events ).
I was able to find an answer on their GitHub https://github.com/zeromq/netmq/issues/845. Seems that the behavior is by design as I got the same with native zmq lib via .NET binding.
I'm building an application that uses TCP sockets to communicate. I want to test how it behaves under slow-speed conditions.
There are similar question on the site, but as I understand it, they deal with HTTP traffic, or are about Linux. My traffic is not HTTP, just ordinary TCP sockets, and the OS is Windows.
I tried using fiddler's setting for Modem Speed but it didn't work, it seems to work only for HTTP connections.
While it is true that you probably want to invest in an extensive set of unit tests, You can simulate various network conditions using VMWare Workstation:
You will have to install a virtual machine for testing, setup bridged networking (for the vm to access your real network) and upload your code to the vm.
After that you can start changing the settings and see how your application performs.
NetLimiter can also be used, but it has fewer options (in your case, packet loss is very interesting to test and is not available in netlimiter).
There is an excellent utility for Windows that can do throttling and much more:
https://jagt.github.io/clumsy/
I think you're taking the wrong approach here.
You can achieve everything that you need with some well designed unit tests. All of the things that a slow network link causes can be simulated in a unit test environment in controlled conditions.
Things that your code MUST handle to deal with "slow" links are just things that you should be dealing with anyway, including:
The correct handling of fragmented messages. All of your network reading code needs to correctly assume that each read will return between 1 byte and the size of your read buffer. You should never assume that you'll get complete 'messages' as TCP knows nothing of your concept of messages.
TCP flow control causing either your synchronous sends to fail with some form of 'try later' error or your async sends to succeed and potentially use an uncontrolled amount of resources (see here for more details). Note that this can happen even on 'fast' links if you are sending faster than the receiver is consuming.
Timeouts - again this isn't limited to "slow" links. All of your timeout handling code should be robust and tested. You may want to make sure that any read timeout is based on any read completing rather than reading a complete message in x time. You may be getting your data at a slow rate but whilst you're still getting data the link is alive.
Connection failure - again not something specific to "slow" links. You need to know how you deal with connections being reset at any time.
In summary nothing you can achieve by running your client and server on a simulated slow network cannot be achieved with a decent set of unit tests and everything that you would want to test on such a link is something that could affect any of your connections on any speed of link.
I'm looking for a mechanism to use to create a simple many-to-many messaging system to allow Windows applications to communicate on a single machine but across sessions and desktops.
I have the following hard requirements:
Must work across all Windows sessions on a single machine.
Must work on Windows XP and later.
No global configuration required.
No central coordinator/broker/server.
Must not require elevated privileges from the applications.
I do not require guaranteed delivery of messages.
I have looked at many, many options. This is my last-ditch request for ideas.
The following have been rejected for violating one or more of the above requirements:
ZeroMQ: In order to do many-to-many messaging a central broker is required.
Named pipes: Requires a central server to receive messages and forward them on.
Multicast sockets: Requires a properly configured network card with a valid IP address, i.e. a global configuration.
Shared Memory Queue: To create shared memory in the global namespace requires elevated privileges.
Multicast sockets so nearly works. What else can anyone suggest? I'd consider anything from pre-packaged libraries to bare-metal Windows API functionality.
(Edit 27 September) A bit more context:
By 'central coordinator/broker/server', I mean a separate process that must be running at the time that an application tries to send a message. The problem I see with this is that it is impossible to guarantee that this process really will be running when it is needed. Typically a Windows service would be used, but there is no way to guarantee that a particular service will always be started before any user has logged in, or to guarantee that it has not been stopped for some reason. Run on demand introduces a delay when the first message is sent while the service starts, and raises issues with privileges.
Multicast sockets nearly worked because it manages to avoid completely the need for a central coordinator process and does not require elevated privileges from the applications sending or receiving multicast packets. But you have to have a configured IP address - you can't do multicast on the loopback interface (even though multicast with TTL=0 on a configured NIC behaves as one would expect of loopback multicast) - and that is the deal-breaker.
Maybe I am completely misunderstanding the problem, especially the "no central broker", but have you considered something based on tuple spaces?
--
After the comments exchange, please consider the following as my "definitive" answer, then:
Use a file-based solution, and host the directory tree on a Ramdisk to insure good performance.
I'd also suggest to have a look at the following StackOverflow discussion (even if it's Java based) for possible pointers to how to manage locking and transactions on the filesystem.
This one (.NET based) may be of help, too.
How about UDP broadcasting?
Couldn't you use a localhost socket ?
/Tony
In the end I decided that one of the hard requirements had to go, as the problem could not be solved in any reasonable way as originally stated.
My final solution is a Windows service running a named pipe server. Any application or service can connect to an instance of the pipe and send messages. Any message received by the server is echoed to all pipe instances.
I really liked p.marino's answer, but in the end it looked like a lot of complexity for what is really a very basic piece of functionality.
The other possibility that appealed to me, though again it fell on the complexity hurdle, was to write a kernel driver to manage the multicasting. There would have been several mechanisms possible in this case, but the overhead of writing a bug-free kernel driver was just too high.
I am going to tell the problem that I have to solve and I need some suggestions if i am in the right path.
The problem is:
I need to create a Windows Service application that receive a request and do some action. (Socket communication) This action is to execute a script (maybe in lua or perl).This script models te bussiness rules of the client, querying in Databases, making request in websites and then send a response to the client.
There are 3 mandatory requirements:
The service will receive a lot of request at the same time. So I think to use the worker's thread model.
The service must have a high throughput. I will have many of requests at the same second.
Low Latency: I must response these requests very quickly.
Every request will generate a log entries. I cant write these log entries in the physical disk at same time the scripts execute because the big I/O time. Probably I will make a queue in memory and others threds will consume this queue and write on disk.
In the future, is possible that two woker's thread have to change messages.
I have to make a protocol to this service. I was thinking to use Thrift, but i don't know the overhead involved. Maybe i will make my own protocol.
To write the windows service, i was thinking in Erlang. Is it a good idea?
Does anyone have suggestions/hints to solve this problem? Which is the better language to write this service?
Yes, Erlang is a good choice if you're know it or ready to learn. With Erlang you don't need any worker thread, just implement your server in Erlang style and you'll receive multithreaded solution automatically.
Not sure how to convert Erlang program to Windows service, but probably it's doable.
Writing to the same log file from many threads are suboptimal because requires locking. It's better to have a log-entries queue (lock-free?) and a separate thread (Erlang process?) that writes them to the file. BTW, are you sure that executing external script in another language is much faster than writing a log-record to the file?
It's doubtfully you'll receive much better performance with your own serialization library than Thrift provides for free. Another option is Google Protocol Buffers, somebody claimed that it's faster.
Theoretically (!) it's possible that Erlang solution won't provide you required performance. In this case consider a compilable language, e.g. C++ and asynchronous networking, e.g. Boost.Asio. But be ready that it's much more complicated than Erlang way.
Question:
Can I use the multiprocessing module together with gevent on Windows in an efficient way?
Scenario:
I have a gevent based Python application doing asynchronous I/O on Windows. The application is mostly I/O bound, but there are spikes of higher CPU load as well. This application would need to control a console application via its stdin and stdout. I cannot modify this console application and the user will be able to use his own custom one, only the text (line) based communication protocol is fixed.
I have a working implementation using subprocess and threads, but I would rather move the whole subprocess based communication code together with those threads into a separate process to turn the main application back to single-threaded. I plan to use the multiprocessing module for this.
Prior reading:
I have been searching the Web a lot and read some source code, so I know that the multiprocessing module is using a Pipe implementation based on named pipes on Windows. A pair of multiprocessing.queue.Queue objects would be used to communicate with the second Python process. These queues are based on that Pipe implementation, e.g. the IPC would be done via named pipes.
The key question is, whether calling the incoming Queue's get method would block gevent's main loop or not. There's a timeout for that method, so I could make it into a loop with a small timeout, but that's not a good solution, since it would still block gevent for small time periods hurting its low I/O latency.
I'm also open to suggestions on how to circumvent the whole problem of using pipes on Windows, which is known to be hard and sometimes fragile. I'm not sure whether shared memory based IPC is possible on Windows or not. Maybe I could wrap the console application in a way which would allow communicating with the child process using network sockets, which is known to work well with gevent.
Please don't question my primary use case, if possible. Thanks.
The Queue's get method is really blocking. Using it with timeout could potentially solve your problem, but it definitely won't be a cleanest solution and, which is the most important, will introduce extra latency for no good reason. Even if it wasn't blocking, that won't be a good solution either. Just because non-blocking itself is not enough, the good asynchronous call/API should smoothly integrate into the I/O framework in use. Be that gevent for Python, libevent for C or Boost ASIO for C++.
The easiest solution would be to use simple I/O by spawning your console applications and attaching to its console in and out descriptors. There are at two major factors to consider:
It will be extremely easy for your clients to write client applications. They will not have to work with any kind of IPC, socket or other code, which could be very hard thing for many. With this approach, application will just read from stdin and write to stdout.
It will be extremely easy to test console applications using this approach as you can manually start them, enter text into console and see results.
Gevent is a perfect fit for async read/write here.
However, the downside is that you will have to start this application, there will be no support for concurrent communication with it, and there will be no support for communication over network. There is even a good example for starters.
To keep it simple but more flexible, you can use TCP/IP sockets. If both client and server are running on the same machine. Also, a good operating system will use IPC as an underlying implementation, so it will be fast. And, if you are worrying about performance of this case, you probably should not use Python at all and look at other technologies.
Even fancies solution – use ZeroC ICE. It is very modern technology allowing almost seamless inter-process communication. It is a CORBA killer, very easy to use. It is heavily used by many, proven to be fastest in its class and rock stable. The beauty of this solution is that you can seamlessly integrate programs in many different languages, like Python, Java, C++ etc. But this will require some of your time to get familiar with a concept. If you decide to go this way, just spend a day reading trough documentation.
Hope it helps. Good luck!
Your question is already quite old. Nevertheless, I would like to recommend http://gehrcke.de/gipc which -- I believe -- would tackle the outlined challenge in a very straight-forward fashion. Basically, it allows you to integrate multiprocessing-based child processes anywhere in your application (also on Windows). Interaction with Process objects (such as calling join()) is gevent-cooperative. Via its pipe management, it allows for cooperatively blocking inter-process communication. However, on Windows, IPC currently is much less efficient than on POSIX-compliant systems (since non-blocking I/O is imitated through a thread pool). Depending on the IPC messaging volume of your application, this might or might not be of significance.