python实现requests重试机制

个人博客导航页（点击右侧链接即可打开个人博客）：大牛带你入门技术栈 

一句话解决方案

import requests
from requests.adapters import HTTPAdapter
 
s = requests.Session()
s.mount('http://', HTTPAdapter(max_retries=3))
s.mount('https://', HTTPAdapter(max_retries=3))
 
s.get('http://example.com', timeout=5)

另外一种方式，深度定制, 但是我尝试了在python2.7的版本下不可以， 因为max_retries传入的参数是整型的， 如果传入Retry对象的话， 会报错：total -= 1 TypeError: unsupported operand type(s) for -=: 'Retry' and 'int'

def retry_test():
    s = requests.Session()

    retries = Retry(total=5,
                    backoff_factor=6,
                    status_forcelist=[500, 502, 503, 504])

    s.mount('http://', HTTPAdapter(max_retries=retries))

    resp = s.get('http://localhost:8088', timeout=3)
    print resp.text

原理

urllib3.connectionpool.py里, urlopen通过递归的方式实现

  def urlopen(
        self,
        method,
        url,
        body=None,
        headers=None,
        retries=None,
        redirect=True,
        assert_same_host=True,
        timeout=_Default,
        pool_timeout=None,
        release_conn=None,
        chunked=False,
        body_pos=None,
        **response_kw
    )
        """remove some codes"""

        try:
            # Request a connection from the queue.
            timeout_obj = self._get_timeout(timeout)
            conn = self._get_conn(timeout=pool_timeout)

            conn.timeout = timeout_obj.connect_timeout

            is_new_proxy_conn = self.proxy is not None and not getattr(
                conn, "sock", None
            )
            if is_new_proxy_conn:
                self._prepare_proxy(conn)

            # Make the request on the httplib connection object.
            httplib_response = self._make_request(
                conn,
                method,
                url,
                timeout=timeout_obj,
                body=body,
                headers=headers,
                chunked=chunked,
            )

            # If we're going to release the connection in ``finally:``, then
            # the response doesn't need to know about the connection. Otherwise
            # it will also try to release it and we'll have a double-release
            # mess.
            response_conn = conn if not release_conn else None

            # Pass method to Response for length checking
            response_kw["request_method"] = method

            # Import httplib's response into our own wrapper object
            response = self.ResponseCls.from_httplib(
                httplib_response,
                pool=self,
                connection=response_conn,
                retries=retries,
                **response_kw
            )

            # Everything went great!
            clean_exit = True

        except queue.Empty:
            # Timed out by queue.
            raise EmptyPoolError(self, "No pool connections are available.")

        except (
            TimeoutError,
            HTTPException,
            SocketError,
            ProtocolError,
            BaseSSLError,
            SSLError,
            CertificateError,
        ) as e:
            # Discard the connection for these exceptions. It will be
            # replaced during the next _get_conn() call.
            clean_exit = False
            if isinstance(e, (BaseSSLError, CertificateError)):
                e = SSLError(e)
            elif isinstance(e, (SocketError, NewConnectionError)) and self.proxy:
                e = ProxyError("Cannot connect to proxy.", e)
            elif isinstance(e, (SocketError, HTTPException)):
                e = ProtocolError("Connection aborted.", e)

            retries = retries.increment(
                method, url, error=e, _pool=self, _stacktrace=sys.exc_info()[2]
            )
            retries.sleep()

            # Keep track of the error for the retry warning.
            err = e

        finally:
            if not clean_exit:
                # We hit some kind of exception, handled or otherwise. We need
                # to throw the connection away unless explicitly told not to.
                # Close the connection, set the variable to None, and make sure
                # we put the None back in the pool to avoid leaking it.
                conn = conn and conn.close()
                release_this_conn = True

            if release_this_conn:
                # Put the connection back to be reused. If the connection is
                # expired then it will be None, which will get replaced with a
                # fresh connection during _get_conn.
                self._put_conn(conn)

        if not conn:
            # Try again
            log.warning(
                "Retrying (%r) after connection broken by '%r': %s", retries, err, url
            )
            return self.urlopen(
                method,
                url,
                body,
                headers,
                retries,
                redirect,
                assert_same_host,
                timeout=timeout,
                pool_timeout=pool_timeout,
                release_conn=release_conn,
                chunked=chunked,
                body_pos=body_pos,
                **response_kw
            )

        def drain_and_release_conn(response):
            try:
                # discard any remaining response body, the connection will be
                # released back to the pool once the entire response is read
                response.read()
            except (
                TimeoutError,
                HTTPException,
                SocketError,
                ProtocolError,
                BaseSSLError,
                SSLError,
            ):
                pass

        # Handle redirect?
        redirect_location = redirect and response.get_redirect_location()
        if redirect_location:
            if response.status == 303:
                method = "GET"

            try:
                retries = retries.increment(method, url, response=response, _pool=self)
            except MaxRetryError:
                if retries.raise_on_redirect:
                    # Drain and release the connection for this response, since
                    # we're not returning it to be released manually.
                    drain_and_release_conn(response)
                    raise
                return response

            # drain and return the connection to the pool before recursing
            drain_and_release_conn(response)

            retries.sleep_for_retry(response)
            log.debug("Redirecting %s -> %s", url, redirect_location)
            return self.urlopen(
                method,
                redirect_location,
                body,
                headers,
                retries=retries,
                redirect=redirect,
                assert_same_host=assert_same_host,
                timeout=timeout,
                pool_timeout=pool_timeout,
                release_conn=release_conn,
                chunked=chunked,
                body_pos=body_pos,
                **response_kw
            )

        # Check if we should retry the HTTP response.
        has_retry_after = bool(response.getheader("Retry-After"))
        if retries.is_retry(method, response.status, has_retry_after):
            try:
                retries = retries.increment(method, url, response=response, _pool=self)
            except MaxRetryError:
                if retries.raise_on_status:
                    # Drain and release the connection for this response, since
                    # we're not returning it to be released manually.
                    drain_and_release_conn(response)
                    raise
                return response

            # drain and return the connection to the pool before recursing
            drain_and_release_conn(response)

            retries.sleep(response)
            log.debug("Retry: %s", url)
            return self.urlopen(
                method,
                url,
                body,
                headers,
                retries=retries,
                redirect=redirect,
                assert_same_host=assert_same_host,
                timeout=timeout,
                pool_timeout=pool_timeout,
                release_conn=release_conn,
                chunked=chunked,
                body_pos=body_pos,
                **response_kw
            )

        return response

参考

https://www.biaodianfu.com/python-requests-retry.html

https://stackoverflow.com/questions/15431044/can-i-set-max-retries-for-requests-request

附Java/C/C++/机器学习/算法与数据结构/前端/安卓/Python/程序员必读/书籍书单大全：

（点击右侧 即可打开个人博客内有干货）：技术干货小栈
=====>>①【Java大牛带你入门到进阶之路】<<====
=====>>②【算法数据结构+acm大牛带你入门到进阶之路】<<===
=====>>③【数据库大牛带你入门到进阶之路】<<=====
=====>>④【Web前端大牛带你入门到进阶之路】<<====
=====>>⑤【机器学习和python大牛带你入门到进阶之路】<<====
=====>>⑥【架构师大牛带你入门到进阶之路】<<=====
=====>>⑦【C++大牛带你入门到进阶之路】<<====
=====>>⑧【ios大牛带你入门到进阶之路】<<====
=====>>⑨【Web安全大牛带你入门到进阶之路】<<=====
=====>>⑩【Linux和操作系统大牛带你入门到进阶之路】<<=====

天下没有不劳而获的果实，望各位年轻的朋友，想学技术的朋友，在决心扎入技术道路的路上披荆斩棘，把书弄懂了，再去敲代码，把原理弄懂了，再去实践，将会带给你的人生，你的工作，你的未来一个美梦。