Hide Forgot
Created attachment 514787 [details] netstate and tcpdump outputs as described above Description of problem: I run a diameter client application that establishes a TCP connection to a diameter server application, it exchanges data, then closes the TCP connection. I run a script the continually executes the application, killing the application after 1 second and then immediately restarting the application. TCP connections are open and closed at a rate of 30 connection/sec. After 2-4 hours, the TCP ports on the server side remain in CLOSE_WAIT indefinitely even after client application is terminated tcp 153 0 hslu19a:diameter hslu18a:54143 CLOSE_WAIT tcp 153 0 hslu19a:diameter hslu18a:54141 CLOSE_WAIT tcp 153 0 hslu19a:diameter hslu18a:54139 CLOSE_WAIT tcp 153 0 hslu19a:diameter hslu18a:54137 CLOSE_WAIT tcp 50961 0 hslu19a:diameter hslu18a:53987 CLOSE_WAIT tcp 153 0 hslu19a:diameter hslu18a:54147 CLOSE_WAIT tcp 153 0 hslu19a:diameter hslu18a:54145 CLOSE_WAIT State of server machine is such that client can no longer establish TCP connection to server Version-Release number of selected component (if applicable): Both client and server machines are running RHEL 3.8 Linux hslu19a 2.4.21-47.ELsmp #1 SMP Wed Jul 5 20:38:41 EDT 2006 i686 i686 i386 GNU/Linux How reproducible: Steps to Reproduce: 1. Run diameter server on server machine 2. On client machine, run script that contimually start client application, kills application after o1 second, then restarts application 3. After some time, observer TCP ports on server in CLOSE_WAIT state with data on RecvQ Actual results: TCP ports on server stuck CLOSE_WAIT state and client unable to establish new TCP connections Expected results: TCP ports should not be stuck in CLOSE_WAIT state Additional info: 1) tcpdump tcp.log3 takea when initial problem occured. Problem starts around Jul 22 10:13:48 2) tcpdump tcpCloseWait.log taken when attempting to open TCP connections after ports stuck in CLOSE_WAIT 2) netstat -a of server ports Addition info avaliable at request. Machines are still in problem state
Thank you for submitting this issue for consideration in Red Hat Enterprise Linux. The release for which you requested us to review is now End of Life. Please See https://access.redhat.com/support/policy/updates/errata/ If you would like Red Hat to re-consider your feature request for an active release, please re-open the request via appropriate support channels and provide additional supporting details about the importance of this issue.