使用HttpClient,總是報出“Going to buffer response body of large or unknown size. Using getResponseBodyAsStream instead is recommended.”的WARN日志,定位到HttpClient的源碼如下:
public abstract class HttpMethodBase implements HttpMethod { ... public byte[] getResponseBody() throws IOException { if (responseBody == null) { InputStream instream = getResponseBodyAsStream(); if (instream != null) { long contentLength = getResponseContentLength(); if (contentLength > 2147483647L){ throw new IOException("Content too large to be buffered: " + contentLength + " bytes");
} int limit = getParams().getIntParameter("http.method.response.buffer.warnlimit", 1048576); if (contentLength == -1L || contentLength > (long) limit){ LOG.warn("Going to buffer response body of large or unknown size. Using getResponseBodyAsStream instead is recommended.");
} LOG.debug("Buffering response body"); ByteArrayOutputStream outstream = new ByteArrayOutputStream(contentLength <= 0L ? 4096: (int) contentLength); byte buffer[] = new byte[4096]; int len; while ((len = instream.read(buffer)) > 0){
outstream.write(buffer, 0, len);
} outstream.close(); setResponseStream(null); responseBody = outstream.toByteArray(); } } return responseBody; } ... }
報WARN的條件是((contentLength == -1) || (contentLength > limit)),也就是說,或者是返回的HTTP頭沒有指定contentLength,或者是contentLength大於上限(默認是1M)。如果能確定返回結果的大小對程序沒有顯著影響,這個WARN就可以忽略,可以在日志的配置中把HttpClient的日志級別調到ERROR,不讓它報出來。
當然,這個警告也是有意義的,HttpClient建議使用InputStream getResponseBodyAsStream()代替byte[] getResponseBody()。對於返回結果很大或無法預知的情況,就需要使用InputStreamgetResponseBodyAsStream(),避免byte[] getResponseBody()可能帶來的內存的耗盡問題。