Table of Contents
linxcfg - a configuration utility for LINX links to other nodes
linxcfg
create <macaddr> <device> <linkname> [opt1=v1 ...] [-v]
linxcfg destroy <linkname>
[-v]
linxcfg -h
linxcfg is the LINX Ethernet configuration command.
- create
- Create an Ethernet connection link to another node with the macaddr
MAC address, through the given device and give it the connection name
linkname to be used as hunt path to reach the other system. Several other
options can be given to configure the connection link. Note that the timeout
values depend on the system clock tick; the configured timeout values are
minimum values.
A remote link cannot be used until also the remote node creates the other
end of the link (a link back to the originating node). The state is "connecting"
until both sides are ready and communication via the LINX protocol can
begin.
- destroy
- Destroy the Ethernet connection link with the name linkname.
- -h
- Display help and exit.
- -v
- Verbose mode.
- macaddr
- Specify the remote mac address
of the node to connect to, e.g. 0a:1b:2c:3d:4d:5e.
- device
- Specify the device
name to use. ex: eth0.
- linkname
- Optional name of the connection link. This
is the LINX link part of the path in a remote hunt name to LINX endpoints
on the other node. It will also be the name seen in /proc/net/linx/cm/eth/<linkname>.
- window_size=nnn
- This is the send/receive window_size, in number of packets
on the link, and may need to be modified to adapt LINX to really slow
or really fast ethernet performance. Both the local and remote side must
use the same window size. The default 128 messages should be sufficient
in most configurations. The window size shall always be of a power of 2.
Size 0 means to use the default window size.
- defer_queue_size=nnn
- The defer
queue size is in packages, with the size depending on the LINX link. The
defer queue is used on the sender side when the send queue is full. Every
message sent, when the send queue is full, is stored in the defer queue,
until the send queue has room for more messages. The defer queue size is
the maximum number of packages, with messages, that can be stored in the
defer queue before the link is disconnected, due to lack or resources.
The default value 2048 packages should be sufficient in most systems. Size
0 means to use the default defer queue size.
- send_tmo=nnn
- The send acknowledge
timeout specifies the time (in msec) to wait until unacknowledged messages
are resent. The default 10 should be sufficient in most systems. Size 0 means
to use the default timeout.
- nack_tmo=nnn
- The retransmission timeout specifies
the time (in msec) to wait until nack messages are resent. The default 20
should be sufficient in most systems. Size 0 means to use the default timeout.
- conn_tmo=nnn
- The connect timeout specifies the time (in msec) to wait until
an attempt to establish a connection is considered failed. The default 200
should be sufficient in most systems. Size 0 means to use the default timeout.
- live_tmo=nnn
- The connection supervision timeout specifies the time (in
msec) to wait until a connection is considered broken. Default is 100 (when
the system is not idle) and should be sufficient in most systems. When the
system is idle, the live timeout is ten times larger than the configured
live timeout value. Size 0 means to use the default timeout.
- coalesce=[his,los,hir,lor
| off]
- NOTE: This feature is still experimental! If this configuration is
set, adaptive software-controlled ethernet coalescing is enabled. The default
is off. The high send/receive thresholds his and hir specify the number
of sent/received messages/200msec that need to be sent/received to enable
coalescing in the specific direction. The low send/receive thresholds los
and lor specify when to disable coalescing in the specific direction. Adaptive
coalesting means to dynamically optimize the throughput by collecting
multiple small messages in each package, but still avoiding delays due
to long time periods between messages. The coalesce configuration may require
device specific tuning, e.g. by running a benchmark with different coalesce
configurations.
None.
linxcfg will display more verbose
information to standard out, if the -v option is specified.
None.
linxcfg create 01:23:a4:4f:b3:ac eth0 conn_A
linxcfg destroy conn_A
linx(7)
, linxstat(1)
, linxdisc(8)
Enea
LINX team
Copyright (c) 2006-2007, Enea Software AB All rights
reserved.
Redistribution and use in source and binary forms, with or without
modification, are permitted provided that the following conditions are
met:
Redistributions of source code must retain the above copyright notice,
this list of conditions and the following disclaimer. Redistributions in
binary form must reproduce the above copyright notice, this list of conditions
and the following disclaimer in the documentation and/or other materials
provided with the distribution. Neither the name of Enea Software AB nor
the names of its contributors may be used to endorse or promote products
derived from this software without specific prior written permission.
THIS
SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS "AS IS"
AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT OWNER OR CONTRIBUTORS BE
LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER
CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY,
OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE
USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
Table of Contents