[ofa-general] RESEND: net.ipv4.tcp_timestamps

Amar Mudrankit amar.mudrankit at qlogic.com
Wed Jan 7 06:15:08 PST 2009


Any updaate / clarification on this? It looks like this is being reset
as mlx4_en tuning parameter.(ofed_scripts/ofa_kernel.spec)

On Mon, Dec 1, 2008 at 9:03 PM, John Russo <john.russo at qlogic.com> wrote:
>
> Does anyone know why this value is being altered by OFED?
>
> "net.ipv4.tcp_timestamps" is being set to 0 during OFED installation.
>
> Default value of this parameter is set to 1 on standard RHEL/SLES distros which OFED installation script modifies to 0.  Also, when OFED is uninstalled, it does not reset these sysctl parameters to their original values.
>
> This parameter is specifically recommended to be turned ON for High performance network.  This is a TCP option that can be used to calculate the Round Trip Measurement in a better and more accurate way than the retransmission timeout method can.  Accurate value of retransmission timeout should be determined to avoid unnecessary retransmissions and hence to improve TCP performance.  RFC 1323 talks about this TCP extension for High Performance.
>
> When the parameter net.ipv4.tcp_timestamps=1, then it adds extra 12 bytes into TCP header increasing its size.  This has an obvious effect of decrease in bandwidth as we have some extra data flowing.  Is this the reason why OFED turns it OFF to net.ipv4.tcp_timestamps=0.
> _______________________________________________
> general mailing list
> general at lists.openfabrics.org
> http://lists.openfabrics.org/cgi-bin/mailman/listinfo/general
>
> To unsubscribe, please visit http://openib.org/mailman/listinfo/openib-general
>



More information about the general mailing list