Take the 2-minute tour ×
Server Fault is a question and answer site for professional system and network administrators. It's 100% free, no registration required.

We have an issue with Windows 2008 Datacenter edition SP2 64bit. We have a process that is polling very frequently and establishing new TCP connections. The system gets in a state where we end up with over 16k connections in TIME_WAIT state. The default OS timeout is 120 seconds after which these connections should go away, but that never happens. These connections persist and never get cleaned up even after the originating process has long terminated (we are still at 16k connections two days after the process was killed). The OS is supposed to time them out but it doesn't.

Has anyone else seen this behavior and if so what was done to resolve it. We are aware of how to tune the tcp stack to make the timeout shorter or allow more connections but this is not the issue here.

Thanks!

share|improve this question

5 Answers 5

Amazon EC2 had a major problem with this. They recently fixed the bug. Maybe the same problem applies in your situation?

Hi, I am pasting below an explanation of what was causing this issue. Good news is that this has been fixed very recently by our engineering team. To get fix, all you'll have to do is STOP/START the Windows Server 2008 instances where you are seeing this issue. Again, I am not talking about REBOOT which is different. STOP/START causes the instance to move to a different (healthy) host. When these instances launch again, they will be running on hosts that have the fix in place so they won't have this issue again. Now below is the engineering explanation of this issue. After an in depth investigation, we've found that when running Windows 2008 x64 on most available instance types, we've identified an issue which may result in TCP connections that remain in TIME_WAIT/CLOSE_WAIT for excessively long periods of time (in some cases, remaining in this state indefinitely). While in these states, the particular socket pairs remain unusable and if enough accumulate, will result in port exhaustion for the ports in question. If this particular circumstance occurs, the only solution to clear the socket pairs in question is to reboot the instance in question. We have determined the cause to be the values produced by a timer function in Windows 2008 kernel API which, on many of our 64-bit platforms, will occasionally retrieve a value that is extremely far in the future. This affects the TCP stack by causing the timestamps on the TCP socket pairs to be stamped significantly far in the future. According to Microsoft, there is a stored cumulative counter which will not be updated unless the value produced by this API call is larger than the cumulative value. The ultimate result is that sockets created after this point will all be stamped too far in the future until that future time is reached. In some cases, we have seen this value several hundred days into the future, thus the socket pairs appear to be stuck forever.

share|improve this answer
1  
Wow great! We have the same problem on EC2, thanks for the solution! –  chris166 Jun 16 '11 at 15:52

There is a Microsoft Article that describes a few ways to resolve this. It commonly comes from Applications that are badly coded and do not close ports correctly. You need to look at what applications you have installed, or what tasks you are performing and disabling these to see which is causing the issue.

To fix the issue, you want to be looking at either;

  1. Increase the upper range of ephemeral ports that are dynamically allocated to client TCP/IP socket connections.
  2. Reduce the client TCP/IP socket connection timeout value from the default value of 240 seconds (A more permanent fix)
share|improve this answer
2  
hyperperforator, that is the common and well-understood manifestation of this problem. The problem that Peco is experiencing (and I as well) is that some time_wait connections are never being released, even after the time_wait delay. Eventually the machine runs out of ports, no matter the rate at which new connections are established. –  user69387 Feb 3 '11 at 18:01

We have same issue, but after connections reach 30k or more, services start failing. Haven't found anything short of reboot that solves this issue.

Running windows 2008 standard on EC2.

share|improve this answer

I had same problem with windows 2003 server. Problem was solved when I reboot the machine after registry TCPIP parameter change.. May be you can give it a try in server 2008

share|improve this answer

I have noticed this problem differs when the same VM (Windows 2008r2) is deployed on either an Intel or an AMD Magny-Cours VMware server. On the AMD the connections stay in TIME_WAIT indefinately, on the Intel machines they obey the standard 4 mins TIME_WAIT timeout.

share|improve this answer

protected by Community Aug 28 '11 at 8:39

Thank you for your interest in this question. Because it has attracted low-quality answers, posting an answer now requires 10 reputation on this site.

Would you like to answer one of these unanswered questions instead?

Not the answer you're looking for? Browse other questions tagged or ask your own question.