Question

I am trying to call a webservice that return too much data just to extract a small piece of data.

So, I decided not to use the standard Client which is generated by Java.

I use the following code to do the connection:

HttpURLConnection connection;
byte[] requestData = .....
URL url = new URL(wsUrl);
connection = (HttpURLConnection) url.openConnection();
connection.setRequestMethod("POST");
connection.setDoOutput(true);
connection.setRequestProperty("Content-Type", "text/xml");
connection.setRequestProperty("Content-Length", String.valueOf(requestData.length));
connection.connect();
OutputStream connOs = connection.getOutputStream();
connOs.write(requestData);
connOs.close(); 
InputStream is = connection.getInputStream();   // <<< THIS IS THE MOST TIME CONSUMING, it takes about 70 ms
byte[] rply = stream2Bytes(is);
is.close();
connection.disconnect();

The most time is consumed in the call to connection.getInputStream(); which it takes about 70ms.

I am trying setting many request headers to reduce this time but cannot reach.

My understanding it that the HttpUrlConnection uses HTTP1.1 protocol that uses Connection=KEEP-ALIVE header by default so that the underlying TCP connection is reused.

Answer 1

connection.getInputStream(); – function which wait for server response… you can’t speed up this proccess.