ChallengeThe race condition in the TCP stack, answer

time to read 3 min | 410 words

In my previous post, I discussed a problem in missing data over TCP connection that happened in a racy manner, only every few hundred runs. As it turns out, there is a simple way to make the code run into the problem every single time.

The full code for the repro can be found here.

Change these lines:

image

And voila, you will consistently run into the problem .  Wait, run that by me again, what is going on here?

As it turns out, the issue is in the server, more specifically, here and here. We use a StreamReader to read the first line from the client, do some processing, and then hand it to the ProcessConnection method, which also uses a StreamReader. More significantly, it uses a different StreamReader.

Why is that significant? Well, because of this, the StreamReader has buffers, by default, that are 1KB in size. So here is what happens in the case above: we send a single packet to the server, and when the first StreamReader reads from the stream, it fills the buffer with the two messages. But since there is a line break between them, when we call ReadLineAsync, we actually only get the first one.

Then, we when get to the ProcessConnection method, we have another StreamReader, which also reads from the stream, but the second message had already been read (and is waiting in the first StreamReader buffer), so we are waiting for more information from the client, which will never come.

So how come it sort of works if we do this in two separate calls? Well, it is all about the speed. In most cases, when we split it into two separate calls, the server socket has only the first message in there when the first StreamReader runs, so the second StreamReader is successful in reading the second line. But in some cases, the client manages being fast enough and sending both messages to the server before the server can read them, and voila, we have the same behavior, only much more unpredictable.

The key problem was that it wasn’t obvious we were reading too much from the stream, and until we figured that one out, we were looking in a completely wrong direction. 

More posts in "Challenge" series:

  1. (19 Jan 2017) What does this code do?
  2. (26 Jul 2016) The race condition in the TCP stack, answer
  3. (25 Jul 2016) The race condition in the TCP stack
  4. (28 Apr 2015) What is the meaning of this change?
  5. (26 Sep 2013) Spot the bug
  6. (27 May 2013) The problem of locking down tasks…
  7. (17 Oct 2011) Minimum number of round trips
  8. (23 Aug 2011) Recent Comments with Future Posts
  9. (02 Aug 2011) Modifying execution approaches
  10. (29 Apr 2011) Stop the leaks
  11. (23 Dec 2010) This code should never hit production
  12. (17 Dec 2010) Your own ThreadLocal
  13. (03 Dec 2010) Querying relative information with RavenDB
  14. (29 Jun 2010) Find the bug
  15. (23 Jun 2010) Dynamically dynamic
  16. (28 Apr 2010) What killed the application?
  17. (19 Mar 2010) What does this code do?
  18. (04 Mar 2010) Robust enumeration over external code
  19. (16 Feb 2010) Premature optimization, and all of that…
  20. (12 Feb 2010) Efficient querying
  21. (10 Feb 2010) Find the resource leak
  22. (21 Oct 2009) Can you spot the bug?
  23. (18 Oct 2009) Why is this wrong?
  24. (17 Oct 2009) Write the check in comment
  25. (15 Sep 2009) NH Prof Exporting Reports
  26. (02 Sep 2009) The lazy loaded inheritance many to one association OR/M conundrum
  27. (01 Sep 2009) Why isn’t select broken?
  28. (06 Aug 2009) Find the bug fixes
  29. (26 May 2009) Find the bug
  30. (14 May 2009) multi threaded test failure
  31. (11 May 2009) The regex that doesn’t match
  32. (24 Mar 2009) probability based selection
  33. (13 Mar 2009) C# Rewriting
  34. (18 Feb 2009) write a self extracting program
  35. (04 Sep 2008) Don't stop with the first DSL abstraction
  36. (02 Aug 2008) What is the problem?
  37. (28 Jul 2008) What does this code do?
  38. (26 Jul 2008) Find the bug fix
  39. (05 Jul 2008) Find the deadlock
  40. (03 Jul 2008) Find the bug
  41. (02 Jul 2008) What is wrong with this code
  42. (05 Jun 2008) why did the tests fail?
  43. (27 May 2008) Striving for better syntax
  44. (13 Apr 2008) calling generics without the generic type
  45. (12 Apr 2008) The directory tree
  46. (24 Mar 2008) Find the version
  47. (21 Jan 2008) Strongly typing weakly typed code
  48. (28 Jun 2007) Windsor Null Object Dependency Facility