- Boost-users - lists.preview.boost.org

[Beast] Client connection times out after 5s. How to avoid?
by Dominique Devienne 15 Dec '20

15 Dec '20

Hi. Continuing my explorations of Boost.Beast. So now I have: * a small Boost.Beast HTTP server (based on async_http_server example), and * a small Boost.Beast HTTP client (based on sync_http_client example). This works, in that I can start the client (wrapped in a helper class), make several requests, and close the client. But until now, the client was short-lived, so no problem. (the server was started too of course, and long lived as expected). Now I'm embedding that HTTP client into another (WebSocket) server, which uses that same wrapper class above, successfully connects to the HTTP server on startup, and now waits for WebSocket connections, at which time it's supposed to make HTTP requests to the HTTP server for various reasons. Problem is, the HTTP client times out after 5s. While I want it to be long lived. I've read Vinnie's answer in https://stackoverflow.com/a/56880415/103724 about sync operation not having timeouts, but I don't want to control timeouts of individual requests here (for now at least), I want to not have the underlying socket not timeout on its own, the same way a WebSocket does not timeout either. Is that possible? Who controls that 5s timeout? Would switching to an async HTTP client make any difference? Sounds to me that it would allow me to control timeouts of requests, but would it make any difference to the socket timeout described above? Thanks, --DD

2 4

[Beast] Slow HTTP connect on Windows. Normal?
by Dominique Devienne 14 Dec '20

14 Dec '20

Hi. I'm writing a small interaction between: 1) a Boost.Beast-based HTTP server (based on http_server_async.cpp example) 2) a Boost.Beast-based HTTP client (based on http_client_sync.cpp example) When I log the time it takes to resolve the server address and connect to it, then issue my HTTP request, it takes over 1 second (Win10, VS2019, C++17, Release, localhost for both client and server): 2020-11-25T09:46:37.473936 Prologue in 0.001s 2020-11-25T09:46:37.482470 Resolved in 0.008s 2020-11-25T09:46:38.506180 Connected in 1.023s 2020-11-25T09:46:38.509607 OK: Authenticated; in 0.003s While the same on Linux (RH7.5) is just over 2ms: 2020-11-25T09:45:45.515926 Prologue in 0.000s 2020-11-25T09:45:45.517083 Resolved in 0.001s 2020-11-25T09:45:45.517550 Connected in 0.000s 2020-11-25T09:45:45.518010 OK: Authenticated; in 0.000s That's a huge difference! Almost 500x... And when I contact the same HTTP server on Windows, but from Chrome this time, it takes about 300ms, and if I hit reload rapidly, the time jumps around to as low as 3ms, and as high as 300ms (the same initial delay), with ~50ms and ~100ms in between. (I also see several different connections being established, for some reason...) I've seen similar differences between Windows and Linux connection times, but with WebSocketPP-based client and server this time, also based on Boost.ASIO. Q1: Am I doing anything wrong? I.e. is this "normal" somehow? Q2: How come Chrome, on Windows too, is 3x faster than ASIO-based clients, with the same Beast-based server? Given the high connect time on Windows, I thought I'd try to keep the connection open on the client, and issue several send-request/read-response pairs, using the Beast-based (sync) HTTP client, but only the first one works correctly, the 2nd errors out with: Error: An established connection was aborted by the software in your host machine I suspect it is a user-error, with the server closing the connection, despite the Keep-Alive HTTP header being present server-side on the request. Q3: Given the Beast HTTP examples I'm based off, would anyone have suggestions on what changes are necessary to allow the client issuing several (non-overlapping) requests to the server, on the same connection? (since connecting is so expensive). Q4: And for better performance, what about overlapping request/response pairs, with HTTP pipelining. How to set that up on the client and server with Beast? Thanks for any help on this. --DD PS: Great examples in Boost.Beast BTW. Thanks Vinnie.

4 9

boost.python and python3 64 bit
by Cruz, Petagay [USA] 14 Dec '20

14 Dec '20

Does Boost Python support using python 3 64 bit? if yes, how do i build say Boost.Python 1.74? What is the 'b2' command and then the bjam command for the tutorial example? Thank you

1 0

Boost Asio support for OpenSSL 1.1.1
by Gehrs, Volker 14 Dec '20

14 Dec '20

Hi all, we are using boost.asio on windows. Does boost.asio support OpenSSL 1.1.1 in any new version? I could not find a hint in the Revision history of boost.asio. https://www.boost.org/doc/libs/1_75_0/doc/html/boost_asio/history.html Thanks in advance!

2 1

unresolved external symbol __DllMainCRTStartup@12
by Cruz, Petagay [USA] 13 Dec '20

13 Dec '20

I have built boost python. I am trying to use the boost_1_74_0\libs\python\example\tutorial example. I keep getting a fail in the build of the dll: msvc.link.dll bin\msvc-14.0\debug\boost_python.dll LINK : warning LNK4001: no object files specified; libraries used LINK : error LNK2001: unresolved external symbol __DllMainCRTStartup@12 bin\msvc-14.0\debug\boost_python.dll : fatal error LNK1120: 1 unresolved externals call "bin\standalone\msvc\msvc-14.0\msvc-setup.bat" x86 >nul link /NOLOGO /INCREMENTAL:NO /DLL /DEBUG /MACHINE:X86 /MANIFEST:EMBED /subsystem:console /out:"bin\msvc-14.0\debug\boost_python.dll" /IMPLIB:"bin\msvc-14.0\debug\boost_python.lib" @"bin\msvc-14.0\debug\boost_python.dll.rsp" Any advice as to what might be wrong. My Visual Studio 2019 does not have the corelibc.lib. I only find that under 'WinCE' libs, and so not found when I compile msvc 14... thanks

1 0

Where to find a complete boost example using ICL (intel 64bit compiler for windows 10)?
by Mehdi Est 13 Dec '20

13 Dec '20

Hi Everyone, Where can I find a complete boost example using ICL (intel 64bit C/C++ compiler for windows)? I would like to do it in header-only mode. Right now, I am compiling and linking using the below-mentioned flags/libraries. However, when I run the code I seem to experience deadlock because the execution stalls. Please let me know your valuable comments. Regards *Compile Flags* icl -I"C:/Program Files (x86)/IntelSWTools/sw_dev_tools/compilers_and_libraries_2020.2.254/windows/compiler/include" -I/../boost_1_73_0 -I../../../boost_1_73_0 -Qstd=c++0x -Qdiag-error-limit:3 -debug:all -Zi -Z7 -ZI -MDd -D_DEBUG -DBOOST_STACKTRACE_USE_WINDBG -Fo:../../Debug/obj/test.obj -c test.cpp *Link Flags* xilink -LIBPATH:"c:\Program Files (x86)\Windows Kits\10\Lib\10.0.19041.0\um\x64" -out:Debug/test.exe test.obj ole32.lib dbgeng.lib

1 0

[release] Boost 1.75.0
by Marshall Clow 11 Dec '20

11 Dec '20

Release 1.54.0 of the Boost C++ Libraries is now available. These open-source libraries work well with the C++ Standard Library, and are usable across a broad spectrum of applications. The Boost license encourages both commercial and non-commercial use. This release contains 1 new library and numerous enhancements and bug fixes for existing libraries. For details, including download links, see <http://www.boost.org/users/news/version_1.75.0> You can also download directly from: <https://dl.bintray.com/boostorg/release/1.75.0/> To install this release on your system, see <http://www.boost.org/doc/libs/release/more/getting_started/index.html> Thanks to everyone who participated in this release. Thanks, -- The Boost Release Team Marshall Clow, Glen Fernandes, Michael Caisse

3 3

Fwd: [Beast] Slow HTTP connect on Windows. Normal?
by Dominique Devienne 11 Dec '20

11 Dec '20

Oops, this was meant for the list of course. --DD ---------- Forwarded message --------- From: Dominique Devienne <ddevienne(a)gmail.com> Date: Fri, Dec 11, 2020 at 6:54 PM Subject: Re: [Boost-users] [Beast] Slow HTTP connect on Windows. Normal? To: Vinnie Falco <vinnie.falco(a)gmail.com> On Wed, Nov 25, 2020 at 7:43 PM Vinnie Falco <vinnie.falco(a)gmail.com> wrote: > On Wed, Nov 25, 2020 at 2:55 AM Dominique Devienne via Boost-users > <boost-users(a)lists.boost.org> wrote: > > Given the high connect time on Windows, I thought I'd try to keep the > connection open on the client, and issue several send-request/read-response > pairs, using the Beast-based (sync) HTTP client, but only the first one > works correctly, the 2nd errors out with: > > > > Error: An established connection was aborted by the software in your > host machine > > This means that the server closed the connection. Are you setting > HTTP/1.1 in the request? For HTTP/1.1 the default is to enable > keep-alive. But for 1.0 the default is to close the connection. > Sorry for the delay. Just getting back to this. Yes, I'm in 1.1, and I verified that the request has Keep-Alive, and I'm (or the example, I don't recall) is setting the same value on the response. > it sounds like your connection is being closed after one request. Beast > definitely supports keep-alive, so this is certainly user error. > Please inspect the request/response messages taking place and post > them on the list unmodified (use a wire sniffer for this if necessary) > if you are still having problems. > Debugging this, I can see that the close arg to on_write is true, despite the Keep-Alive on the response. That response has an empty_body BTW, in case it matters. This route just replies 200 OK or 403 Forbidden, there's no body to return, only the headers matter to the client. The reason why the connection is closed after the 1st request seems to be because of need_eof(). The async server example has something like this in `send()`: template <typename RES> void send(RES&& msg) { // The lifetime of the message has to extend // for the duration of the async operation so // we use a shared_ptr to manage it. auto sp = std::make_shared<RES>(std::move(msg)); // Store a type-erased version of the shared // pointer in the class to keep it alive. res_ = sp; bool need_eof = sp->need_eof(); ctx_.out() << "send: sp->need_eof() = " << need_eof << std::endl; // Write the response http::async_write( stream_, *sp, beast::bind_front_handler( &session::on_write, shared_from_this(), sp->need_eof() ) ); } And `need_eof()` is true indeed. Which thus calls `on_write()` with `bool close = true`, resulting in a `do_close()` instead of waiting for the next request: void on_write( bool close, beast::error_code ec, std::size_t bytes_transferred ) { boost::ignore_unused(bytes_transferred); if (ec) { return; // fail(ec, "write"); } if (close) { // This means we should close the connection, usually because // the response indicated the "Connection: close" semantic. ctx_.out() << "on_write: close()" << std::endl; return do_close(); } // We're done with the response so delete it res_.reset(); // Read another request do_read(); } I stepped into `need_eof()`, and all conditions of the `if` are false, resulting in the `return true` at the end. template<bool isRequest, class Body, class Fields> bool message<isRequest, Body, Fields>:: need_eof(std::false_type) const { // VFALCO Do we need a way to let the caller say "the body is intentionally skipped"? if( this->result() == status::no_content || this->result() == status::not_modified || to_status_class(this->result()) == status_class::informational || has_content_length() || chunked()) return ! keep_alive(); return true; } So is the issue having an empty_body? I'm not supposed to have one on a GET? BTW, I tried commenting out the `do_close()` inside `on_write()`, just to see what happens, and it still doesn't work, with the same error, but an additional delay (timeout?) I think. Here's the response I'm sending: ctx_.out() << "Sending response: 200 OK" << std::endl; http::response<http::empty_body> res{http::status::ok, req_.version()}; res.set(http::field::server, BOOST_BEAST_VERSION_STRING); res.keep_alive(req_.keep_alive()); ctx_.out() << "on_read: res.keep_alive() = " << res.keep_alive() << std::endl; return send(std::move(res)); I'm fairly sure it's user-error, but it's not easy to figure out for a Beast novice like myself. --DD

2 2

Reminder that Outcome v2.2 is to become Boost.Outcome
by Niall Douglas 11 Dec '20

11 Dec '20

After three Boost releases and a year of warning, this is a reminder that Outcome v2.2 is to replace v2.1 as Boost.Outcome. This is a breaking change, all existing Boost.Outcome code shall need to be upgraded as of Boost 1.76 onwards. Changes: https://www.boost.org/doc/libs/1_75_0/libs/outcome/doc/html/changelog/v22.h… Upgrade guide: https://www.boost.org/doc/libs/1_75_0/libs/outcome/doc/html/changelog/upgra… You should see the merge land into boostorg/outcome next few days. It'll take some weeks to upgrade the tutorial and documentation. Niall

1 0

Boost.Beast base64 decode
by Dominique Devienne 11 Dec '20

11 Dec '20

I'm decoding a Java Web Token, composed of 3 separate base64 encoded parts (header.payload.signature). Before I was using Boost.Serialization: using namespace boost::archive::iterators; using It = transform_width<binary_from_base64<std::string::const_iterator>, 8, 6>; std::string tmp(It(std::begin(base64_str)), It(std::end(base64_str))); return boost::algorithm::trim_right_copy_if(tmp, [](char c) { return c == '\0'; }); But since I'm already using Boost.Beast, I thought I'd drop the above dependency, and use this instead: using namespace boost::beast::detail; std::string decoded(base64::decoded_size(base64_str.size()), '\0'); auto rc = base64::decode(decoded.data(), base64_str.data(), base64_str.size()); decoded.resize(rc.first); But turns out some (all?) JWTs do not include the trailing equals that "proper" base64 should have. And in that case, Beast's rc.first does not include the last character, which I can still see in decoded variable prior to the .resize() in the debugger. Here's the payload part: eyJwYXNzd29yZCI6IlAiLCJ1c2VybmFtZSI6IlUifQ What it should decode to: {"password":"P","username":"U"} What Beast's decode yields: {"password":"P","username":"U" Is this on purpose? Could Beast's base64::decode() be changed to be more lenient about the absence of trailing equals from the base64 encoded string? JWT handling seems like a natural fit within the context of Beast, and although this is an implementation detail, it would seem logical for decode() to cope with that, no? Thanks, --DD

3 4