libhv http client vs cpr

libhv http client 和 cpr 的性能对比

libhv

test code

static void test_http_async(HttpClient* cli, int seq, int* resp_cnt) 
{auto req = std::make_shared<HttpRequest>();req->method = HTTP_GET;req->url = "www.baidu.com";req->timeout = 10;cli->sendAsync(req, [seq, resp_cnt](const HttpResponsePtr& resp) {std::string path = "async-result-" + std::to_string(seq) + ".log";std::ofstream fout(path.c_str());fout << "test_http_async_client response thread" << std::endl;fout << "tid=" << hv_gettid() << std::endl;if (resp == NULL) {fout << "request failed!" << std::endl;} else {fout << resp->status_code << "\t" <<  resp->status_message() << std::endl;// fout << resp->body << std::endl;}*resp_cnt += 1;});
}static void test_http_sync(HttpClient* cli, int seq) 
{HttpRequest req;req.method = HTTP_POST;req.url = "www.baidu.com";req.timeout = 10;HttpResponse resp;int ret = cli->send(&req, &resp);std::string path = "sync-result-" + std::to_string(seq) + ".log";std::ofstream fout(path.c_str());fout << "test_http_sync_client response thread" << std::endl;if (ret != 0) {fout << "request failed!" << std::endl;}else {fout << resp.status_code << "\t" <<  resp.status_message() << std::endl;// fout << resp->body << std::endl;}
}int main(int argc, char* argv[]) {int req_cnt = 0;if (argc > 1) req_cnt = atoi(argv[1]);if (req_cnt == 0) req_cnt = 1;HttpClient sync_client;HttpClient async_client;int resp_cnt = 0;long long beg = gettimeofday_ms();printf("test_http_async_client request thread tid=%ld\n", hv_gettid());for (int i = 0; i < req_cnt; ++i) {test_http_async(&async_client, i, &resp_cnt);}// demo wait async finishedwhile (resp_cnt < req_cnt) hv_delay(10);long long used = gettimeofday_ms() - beg;std::cout << "async finished: " << used << std::endl;beg = gettimeofday_ms();printf("test_http_sync_client request thread tid=%ld\n", hv_gettid());for (int i = 0; i < req_cnt; ++i) {test_http_sync(&async_client, i);}used = gettimeofday_ms() - beg;std::cout << "sync finished: " << used << std::endl;return 0;
}

result

test_http_async_client request thread tid=104820
async finished: 3249ms
test_http_sync_client request thread tid=104820
sync finished: 80398ms[wei@localmachine /home/wei/myworks/libhv]
$cat async-result-*.log | grep "tid=" | sort | uniq -c1000 tid=104823

cpr

test code


void test_http_async(int seq, int *resp_cnt, vector<std::shared_future<string> >& futures)
{auto callback = [seq, resp_cnt](cpr::Response r) {std::string path = "async-result-" + std::to_string(seq) + ".log";std::ofstream fout(path.c_str());fout << "test async response thread" << std::endl;fout << "tid=" << std::this_thread::get_id() << std::endl;fout << r.status_code << "\t" <<  r.status_line << std::endl;if (r.status_code != 200){fout << "error:" << r.error.message << endl;}// else// {//     fout << r.text << std::endl;// }*resp_cnt += 1;return r.text;};string url = "http://www.baidu.com";int timeout = 10 * 1000;futures[seq] = cpr::GetCallback(callback, cpr::Url{url}, cpr::Timeout{timeout});
}void test_http_sync(int seq)
{string url = "http://www.baidu.com";int timeout = 10 * 1000;auto r = cpr::Get(cpr::Url{url}, cpr::Timeout{timeout});std::string path = "sync-result-" + std::to_string(seq) + ".log";std::ofstream fout(path.c_str());fout << "test sync response thread" << std::endl;fout << "tid=" << std::this_thread::get_id() << std::endl;fout << r.status_code << "\t" <<  r.status_line << std::endl;if (r.status_code != 200){fout << "error:" << r.error.message << endl;}// else// {//     fout << r.text << std::endl;// }
}void CprRequestTest::AsyncSyncRequestTest()
{int req_cnt = 1000;int resp_cnt = 0;vector<std::shared_future<string>> cprfutures(req_cnt);cprfutures.reserve(1000);int64_t beg = Timer::GetCurrentTimeInMilliSeconds();cout << "test async request thread tid=" << std::this_thread::get_id() << endl;;for (int i = 0; i < req_cnt; ++i) {test_http_async(i, &resp_cnt, cprfutures);// cprfutures.emplace_back(f);}// demo wait async finishedwhile (resp_cnt < req_cnt) usleep(10 * 1000);int64_t used = Timer::GetCurrentTimeInMilliSeconds() - beg;std::cout << "async finished: " << used << "ms" << std::endl;beg = Timer::GetCurrentTimeInMilliSeconds();cout << "test sync request thread tid=" << std::this_thread::get_id() << endl;;for (int i = 0; i < req_cnt; ++i) {test_http_sync(i);}used = Timer::GetCurrentTimeInMilliSeconds() - beg;std::cout << "sync finished: " << used << "ms" << std::endl;
}

result

test async request thread tid=140052864329472
async finished: 554ms
test sync request thread tid=140052864329472
sync finished: 32382ms[wei@localmachine /home/wei/myworks/cpr]
cat async-result-*.log | grep "tid=" | sort | uniq -c | wc -l
1000

附

机器

processor       : 0
vendor_id       : GenuineIntel
cpu family      : 6
model           : 63
model name      : Intel(R) Xeon(R) CPU E5-2640 v3 @ 2.60GHz
stepping        : 2
microcode       : 0x3a
cpu MHz         : 2600.000
cache size      : 20480 KB
physical id     : 0
siblings        : 16
core id         : 0
cpu cores       : 16
apicid          : 10
initial apicid  : 10
fpu             : yes
fpu_exception   : yes
cpuid level     : 15
wp              : yes
flags           : fpu vme de pse tsc msr pae mce cx8 apic sep mtrr pge mca cmov pat pse36 clflush dts acpi mmx fxsr sse sse2 ss ht tm pbe syscall nx pdpe1gb rdtscp lm constant_tsc arch_perfmon pebs bts rep_good nopl xtopology nonstop_tsc aperfmperf eagerfpu pni pclmulqdq dtes64 ds_cpl vmx smx est tm2 ssse3 fma cx16 xtpr pdcm pcid dca sse4_1 sse4_2 x2apic movbe popcnt tsc_deadline_timer aes xsave avx f16c rdrand lahf_lm abm arat epb pln pts dtherm tpr_shadow vnmi flexpriority ept vpid fsgsbase tsc_adjust bmi1 avx2 smep bmi2 erms invpcid cqm xsaveopt cqm_llc cqm_occup_llc
bogomips        : 5187.90
clflush size    : 64
cache_alignment : 64
address sizes   : 46 bits physical, 48 bits virtual
power management:[wei@localmachine /home/wei]
cat /proc/cpuinfo | grep processor | wc -l
16