Description of problem: The current RHEL5 sctp implementation has a problem in which an associations receive memory can be exhausted prior to its rwnd closing. I've fixed it upstream and it needs to be backported
This request was evaluated by Red Hat Product Management for inclusion in a Red Hat Enterprise Linux maintenance release. Product Management has requested further review of this request by Red Hat Engineering, for potential inclusion in a Red Hat Enterprise Linux Update release for currently deployed products. This request is not yet committed for inclusion in an Update release.
*** Bug 210208 has been marked as a duplicate of this bug. ***
Since Bug 210208 was closed as a duplicate, I am moving all of the RHEL 5.2 request settings to this BZ.
in 2.6.18-58.el5 You can download this test kernel from http://people.redhat.com/dzickus/el5
Ran a quick iperftp test on CPI1 installed by Patrick. Linux stitch-1-2.testing.local 2.6.18-58.el5PAE #1 SMP Tue Nov 27 17:15:24 EST 2007 i686 i686 i386 GNU/Linux 1 bytes : [ 3] 3.0- 6.0 sec 0.09 MBytes 0.25 Mbits/sec 2 bytes : [ 3] 3.0- 6.0 sec 0.18 MBytes 0.51 Mbits/sec 4 bytes : [ 3] 3.0- 6.0 sec 0.32 MBytes 0.89 Mbits/sec 8 bytes : [ 3] 3.0- 6.0 sec 0.74 MBytes 2.07 Mbits/sec 16 bytes : [ 3] 3.0- 6.0 sec 1.47 MBytes 4.10 Mbits/sec 32 bytes : [ 3] 3.0- 6.0 sec 2.44 MBytes 6.82 Mbits/sec 64 bytes : [ 3] 3.0- 6.0 sec 5.01 MBytes 14.0 Mbits/sec 128 bytes : [ 3] 3.0- 6.0 sec 11.7 MBytes 32.6 Mbits/sec 256 bytes : [ 3] 3.0- 6.0 sec 23.0 MBytes 64.3 Mbits/sec 512 bytes : [ 3] 3.0- 6.0 sec 44.7 MBytes 125 Mbits/sec 1024 bytes : [ 3] 3.0- 6.0 sec 89.9 MBytes 251 Mbits/sec 2048 bytes : [ 3] 3.0- 6.0 sec 172 MBytes 480 Mbits/sec 4096 bytes : [ 3] 3.0- 6.0 sec 316 MBytes 883 Mbits/sec 8192 bytes : [ 3] 3.0- 6.0 sec 529 MBytes 1480 Mbits/sec 16384 bytes : [ 3] 3.0- 6.0 sec 802 MBytes 2242 Mbits/sec 32768 bytes : [ 3] 3.0- 6.0 sec 2.22 MBytes 6.20 Mbits/sec 65536 bytes : [ 3] 3.0- 6.0 sec 0.50 MBytes 1.40 Mbits/sec This time I'm getting some problems with 32K messages also, which makes no sense to me as the default receive buffer seems to be bigger (in /proc/sys/net/sctp/sctp_rmem) than it used to be. Internal Status set to 'Waiting on Support' Status set to: Waiting on Tech This event sent from IssueTracker by streeter issue 103872
An advisory has been issued which should help the problem described in this bug report. This report is therefore being closed with a resolution of ERRATA. For more information on the solution and/or where to find the updated files, please follow the link below. You may reopen this bug report if the solution does not work for you. http://rhn.redhat.com/errata/RHBA-2008-0314.html