Question about ClientAliveCountMax

Jan Pechanec jan.pechanec at oracle.com
Fri Nov 8 07:38:12 EST 2013


On Thu, 7 Nov 2013, Vincenzo Romano wrote:

>>>>>Hi all.
>>>>>What is the expected/intended semantics of setting ClientAliveCountMax
>>>>>to 0 and ClientAliveInterval to non-zero?
>>>>
>>>>         I believe it should work as an idle timeout from the server side
>>>> to disconnect idle clients.  J.
>>>
>>>I think you misunderstood me.
>>
>>         I didn't, you asked what would happen and I told you.  I take it
>> you haven't try that.
>>
>>         /* timeout, check to see how many we have had */
>>         if (packet_inc_alive_timeouts() > options.client_alive_count_max) {
>>                 logit("Timeout, client not responding.");
>>                 cleanup_exit(255);
>>         }
>>
>>>ClientAliveCountMax is a counter. ClientAliveInterval is a timeout interval.
>
>With ClientAliveCountMax == 0 there will be no "client alive packet"
>sent and I will force a disconnection if there is no traffic within
>ClientAliveInterval secs.

	yes, that's how I remember it and that's why it can be 
"overloaded" and used to disconnect idle clients since nothing is 
actually sent from the server.

>With ClientAliveCountMax == 1 there will be 1 "client alive packet"
>sent and I will force a disconnection if there is no traffic and no
>reply within 2*ClientAliveInterval secs.

	yes, I believe that is correct.  You can easily verify that if 
you set up a firewall, for example.

	cheers, J.

>
>Is this correct?
>Sorry for asking silly questions and abusing your patience!
>_______________________________________________
>openssh-unix-dev mailing list
>openssh-unix-dev at mindrot.org
>https://lists.mindrot.org/mailman/listinfo/openssh-unix-dev
>

-- 
Jan Pechanec <jan.pechanec at oracle.com>


More information about the openssh-unix-dev mailing list