Differences between version 3 and predecessor to the previous major change of SysControls.
Other diffs: Previous Revision, Previous Author, or view the Annotated Edit History
| Newer page: | version 3 | Last edited on Tuesday, October 12, 2004 1:50:41 pm | by JohnMcPherson | Revert |
| Older page: | version 1 | Last edited on Wednesday, September 15, 2004 4:46:40 pm | by PerryLorier | Revert |
@@ -1,23 +1,53 @@
This page describes the various linux syscontrols and what they do, these can be viewed/edited via the sysctl(8) command, or via /proc/sys/. eg, kernel.panic can be viewed by
- kernel.panic
+ /sbin/sysctl
kernel.panic
or
cat /proc/sys/kernel/panic
!!!kernel
!!kernel.panic
-This is the number of seconds to wait after a kernel has paniced
before the machine will reboot itself automatically. Very useful for unattended servers, or machines that are difficult to get physical access to.
+This is the number of seconds to wait after a kernel has panicked
before the machine will reboot itself automatically. Very useful for unattended servers, or machines that are difficult to get physical access to.
!!!net
-These sysctl's
control the networking. See also http://bec.at/support/ipsysctl-tutorial/tcpvariables.html
+These sysctls
control the networking. See also http://bec.at/support/ipsysctl-tutorial/tcpvariables.html
!!net.ipv4
These control most of the IPv4 options
!net.ipv4.ip_forwarding
enable global IP forwarding. Very imporant.
!net.ipv4.tcp_fack
Enable __F__orward __Ack__nowledgement congestion avoidance and fast retransmission. The value is not used, if net.ipv4.tcp_sack is not enabled. See http://www.psc.edu/networking/papers/fack_abstract.html for details on how this option works. Basically it seems to assume that missing sequence ranges are dropped (ie, implies no reordering). Linux will disable fack on a per connection basis if it detects reordering.
+
+!net.ipv4.tcp_vegas_cong_avoid (0)
+Enable tcp vegas congestion avoidance. TCP Vegas is a congestion control algorithm that mostly these days is not considered to be particularly useful (see TCP Westward, below). TCP Vegas is a sender side (ie only used by the machine initiating the TCP connection) change and causes TCP to back off when it detects the [RTT] changing (ie, queuing is occuring). It works well when there is only a few flows using the bottleneck link.
+
+!net.ipv4.tcp_westwood (0)
+Enable TCP Westwood+ congestion control algorithm. This is a sender side change (like TCP Vegas) and estimates of throughput are kept to try and make sure that the stack uses the optimum amount of bandwidth at all times. Very useful, defaults off should probably be enabled for many sites.
+
+!net.ipv4.tcp_low_latency (0)
+"If set, the TCP stack makes decisions that prefer lower latency as opposed to higher throughput" This seems to disable a function called tcp_prequeue, no idea what it does.
+
+!net.ipv4.tcp_reordering (3)
+How many duplicate acks you need before you enter fast retransmit. IF you are on a network with lots of reordering then this will need to be raised, linux can dynamically tune this on a per TCP flow, so changing it is normally not that necessary.
+
+!net.ipv4.tcp_sack (1)
+Enable Selective Acknowledgement, reduces the number of segments that need to be retransmitted when packet loss occurs. Good to have on, needed to be enabled by both the sender and the reciever.
+
+!net.ipv4.tcp_ecn (0)
+Explicit Congestion Notification, this can be used by routers on the internet to signal that congestion is imminent and to therefore to slow down sending before packet loss actually occurs. However many firewalls on the internet incorrectly detect the [ECN] data as an attack and drop all packets using ECN, sigh.
+
+!net.ipv4.tcp_retries1 (3)
+How many times to send a [SYN]/[ACK] packet before giving up on a connection.
+
+!net.ipv4.tcp_retries2 (15)
+How many times to send a [TCP] data packet before giving up on a connection.
+
+!inet.ipv4.tcp_syn_retrans (5)
+How many times to send a [SYN] packet before giving up on a connection.
+
+!inet.ipv4.tcp_retrans_collapse (1)
+If retransmissions should be sent as full sized packets working around some [TCP] implementation bugs (?).
----
