Re: new sysctl tunable knob for tcpip

2017-08-01 Thread Massimo Sala
Ok, I understand the lesson. Now the practice : I want to set a 200 seconds timeout, valid on all the interfaces, on all the connections. How can I calculate the value to set tcp_retries2 to ? best regards, Sala On 31/07/2017, valdis.kletni...@vt.edu wrote: > On Mon,

Re: new sysctl tunable knob for tcpip

2017-07-31 Thread valdis . kletnieks
On Mon, 31 Jul 2017 15:16:34 +0200, Massimo Sala said: > I wish to suggest to developers to add this new knob : Note that most of the existing knobs were chosen fairly carefully, and that sometimes, the values chosen aren't immediately obvious, because they have to also take into account

Re: new sysctl tunable knob for tcpip

2017-07-31 Thread Massimo Sala
I wish to suggest to developers to add this new knob : tcp_retries2_time - INTEGER This value influences the timeout (seconds) of an alive TCP connection, when RTO retransmissions remain unacknowledged. RFC 1122 recommends at least 100 seconds for the timeout,

Re: new sysctl tunable knob for tcpip

2017-07-18 Thread valdis . kletnieks
On Thu, 06 Jul 2017 09:23:46 +0200, Massimo Sala said: > I have an idea about a new sysctl knob. It is under net.ipv4. Step 0: Identify whether it is even a good idea. TCP/IP is tougher than you think, especially when you get into congestion control. Step 0.1: Figure out if your brilliant idea

Re: new sysctl tunable knob for tcpip

2017-07-17 Thread Greg KH
On Thu, Jul 06, 2017 at 09:23:46AM +0200, Massimo Sala wrote: > I have an idea about a new sysctl knob. It is under net.ipv4. > > q1) which is the right kernel m-list to discuss it ? net...@vger.kernel.org > 2) any suggestion how to gain the critical mass of consensus ? > > 3) at first : is it