^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1) .. SPDX-License-Identifier: GPL-2.0
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 2)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 3) ==========================================
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 4) EQL Driver: Serial IP Load Balancing HOWTO
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 5) ==========================================
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 6)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 7) Simon "Guru Aleph-Null" Janes, simon@ncm.com
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 8)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 9) v1.1, February 27, 1995
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 10)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 11) This is the manual for the EQL device driver. EQL is a software device
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 12) that lets you load-balance IP serial links (SLIP or uncompressed PPP)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 13) to increase your bandwidth. It will not reduce your latency (i.e. ping
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 14) times) except in the case where you already have lots of traffic on
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 15) your link, in which it will help them out. This driver has been tested
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 16) with the 1.1.75 kernel, and is known to have patched cleanly with
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 17) 1.1.86. Some testing with 1.1.92 has been done with the v1.1 patch
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 18) which was only created to patch cleanly in the very latest kernel
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 19) source trees. (Yes, it worked fine.)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 20)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 21) 1. Introduction
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 22) ===============
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 23)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 24) Which is worse? A huge fee for a 56K leased line or two phone lines?
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 25) It's probably the former. If you find yourself craving more bandwidth,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 26) and have a ISP that is flexible, it is now possible to bind modems
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 27) together to work as one point-to-point link to increase your
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 28) bandwidth. All without having to have a special black box on either
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 29) side.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 30)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 31)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 32) The eql driver has only been tested with the Livingston PortMaster-2e
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 33) terminal server. I do not know if other terminal servers support load-
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 34) balancing, but I do know that the PortMaster does it, and does it
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 35) almost as well as the eql driver seems to do it (-- Unfortunately, in
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 36) my testing so far, the Livingston PortMaster 2e's load-balancing is a
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 37) good 1 to 2 KB/s slower than the test machine working with a 28.8 Kbps
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 38) and 14.4 Kbps connection. However, I am not sure that it really is
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 39) the PortMaster, or if it's Linux's TCP drivers. I'm told that Linux's
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 40) TCP implementation is pretty fast though.--)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 41)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 42)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 43) I suggest to ISPs out there that it would probably be fair to charge
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 44) a load-balancing client 75% of the cost of the second line and 50% of
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 45) the cost of the third line etc...
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 46)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 47)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 48) Hey, we can all dream you know...
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 49)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 50)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 51) 2. Kernel Configuration
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 52) =======================
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 53)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 54) Here I describe the general steps of getting a kernel up and working
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 55) with the eql driver. From patching, building, to installing.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 56)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 57)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 58) 2.1. Patching The Kernel
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 59) ------------------------
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 60)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 61) If you do not have or cannot get a copy of the kernel with the eql
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 62) driver folded into it, get your copy of the driver from
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 63) ftp://slaughter.ncm.com/pub/Linux/LOAD_BALANCING/eql-1.1.tar.gz.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 64) Unpack this archive someplace obvious like /usr/local/src/. It will
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 65) create the following files::
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 66)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 67) -rw-r--r-- guru/ncm 198 Jan 19 18:53 1995 eql-1.1/NO-WARRANTY
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 68) -rw-r--r-- guru/ncm 30620 Feb 27 21:40 1995 eql-1.1/eql-1.1.patch
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 69) -rwxr-xr-x guru/ncm 16111 Jan 12 22:29 1995 eql-1.1/eql_enslave
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 70) -rw-r--r-- guru/ncm 2195 Jan 10 21:48 1995 eql-1.1/eql_enslave.c
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 71)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 72) Unpack a recent kernel (something after 1.1.92) someplace convenient
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 73) like say /usr/src/linux-1.1.92.eql. Use symbolic links to point
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 74) /usr/src/linux to this development directory.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 75)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 76)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 77) Apply the patch by running the commands::
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 78)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 79) cd /usr/src
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 80) patch </usr/local/src/eql-1.1/eql-1.1.patch
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 81)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 82)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 83) 2.2. Building The Kernel
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 84) ------------------------
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 85)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 86) After patching the kernel, run make config and configure the kernel
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 87) for your hardware.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 88)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 89)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 90) After configuration, make and install according to your habit.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 91)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 92)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 93) 3. Network Configuration
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 94) ========================
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 95)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 96) So far, I have only used the eql device with the DSLIP SLIP connection
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 97) manager by Matt Dillon (-- "The man who sold his soul to code so much
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 98) so quickly."--) . How you configure it for other "connection"
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 99) managers is up to you. Most other connection managers that I've seen
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 100) don't do a very good job when it comes to handling more than one
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 101) connection.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 102)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 103)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 104) 3.1. /etc/rc.d/rc.inet1
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 105) -----------------------
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 106)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 107) In rc.inet1, ifconfig the eql device to the IP address you usually use
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 108) for your machine, and the MTU you prefer for your SLIP lines. One
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 109) could argue that MTU should be roughly half the usual size for two
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 110) modems, one-third for three, one-fourth for four, etc... But going
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 111) too far below 296 is probably overkill. Here is an example ifconfig
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 112) command that sets up the eql device::
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 113)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 114) ifconfig eql 198.67.33.239 mtu 1006
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 115)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 116) Once the eql device is up and running, add a static default route to
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 117) it in the routing table using the cool new route syntax that makes
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 118) life so much easier::
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 119)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 120) route add default eql
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 121)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 122)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 123) 3.2. Enslaving Devices By Hand
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 124) ------------------------------
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 125)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 126) Enslaving devices by hand requires two utility programs: eql_enslave
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 127) and eql_emancipate (-- eql_emancipate hasn't been written because when
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 128) an enslaved device "dies", it is automatically taken out of the queue.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 129) I haven't found a good reason to write it yet... other than for
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 130) completeness, but that isn't a good motivator is it?--)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 131)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 132)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 133) The syntax for enslaving a device is "eql_enslave <master-name>
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 134) <slave-name> <estimated-bps>". Here are some example enslavings::
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 135)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 136) eql_enslave eql sl0 28800
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 137) eql_enslave eql ppp0 14400
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 138) eql_enslave eql sl1 57600
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 139)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 140) When you want to free a device from its life of slavery, you can
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 141) either down the device with ifconfig (eql will automatically bury the
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 142) dead slave and remove it from its queue) or use eql_emancipate to free
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 143) it. (-- Or just ifconfig it down, and the eql driver will take it out
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 144) for you.--)::
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 145)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 146) eql_emancipate eql sl0
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 147) eql_emancipate eql ppp0
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 148) eql_emancipate eql sl1
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 149)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 150)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 151) 3.3. DSLIP Configuration for the eql Device
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 152) -------------------------------------------
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 153)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 154) The general idea is to bring up and keep up as many SLIP connections
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 155) as you need, automatically.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 156)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 157)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 158) 3.3.1. /etc/slip/runslip.conf
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 159) ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 160)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 161) Here is an example runslip.conf::
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 162)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 163) name sl-line-1
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 164) enabled
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 165) baud 38400
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 166) mtu 576
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 167) ducmd -e /etc/slip/dialout/cua2-288.xp -t 9
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 168) command eql_enslave eql $interface 28800
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 169) address 198.67.33.239
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 170) line /dev/cua2
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 171)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 172) name sl-line-2
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 173) enabled
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 174) baud 38400
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 175) mtu 576
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 176) ducmd -e /etc/slip/dialout/cua3-288.xp -t 9
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 177) command eql_enslave eql $interface 28800
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 178) address 198.67.33.239
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 179) line /dev/cua3
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 180)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 181)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 182) 3.4. Using PPP and the eql Device
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 183) ---------------------------------
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 184)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 185) I have not yet done any load-balancing testing for PPP devices, mainly
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 186) because I don't have a PPP-connection manager like SLIP has with
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 187) DSLIP. I did find a good tip from LinuxNET:Billy for PPP performance:
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 188) make sure you have asyncmap set to something so that control
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 189) characters are not escaped.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 190)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 191)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 192) I tried to fix up a PPP script/system for redialing lost PPP
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 193) connections for use with the eql driver the weekend of Feb 25-26 '95
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 194) (Hereafter known as the 8-hour PPP Hate Festival). Perhaps later this
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 195) year.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 196)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 197)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 198) 4. About the Slave Scheduler Algorithm
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 199) ======================================
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 200)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 201) The slave scheduler probably could be replaced with a dozen other
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 202) things and push traffic much faster. The formula in the current set
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 203) up of the driver was tuned to handle slaves with wildly different
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 204) bits-per-second "priorities".
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 205)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 206)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 207) All testing I have done was with two 28.8 V.FC modems, one connecting
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 208) at 28800 bps or slower, and the other connecting at 14400 bps all the
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 209) time.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 210)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 211)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 212) One version of the scheduler was able to push 5.3 K/s through the
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 213) 28800 and 14400 connections, but when the priorities on the links were
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 214) very wide apart (57600 vs. 14400) the "faster" modem received all
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 215) traffic and the "slower" modem starved.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 216)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 217)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 218) 5. Testers' Reports
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 219) ===================
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 220)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 221) Some people have experimented with the eql device with newer
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 222) kernels (than 1.1.75). I have since updated the driver to patch
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 223) cleanly in newer kernels because of the removal of the old "slave-
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 224) balancing" driver config option.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 225)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 226)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 227) - icee from LinuxNET patched 1.1.86 without any rejects and was able
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 228) to boot the kernel and enslave a couple of ISDN PPP links.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 229)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 230) 5.1. Randolph Bentson's Test Report
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 231) -----------------------------------
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 232)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 233) ::
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 234)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 235) From bentson@grieg.seaslug.org Wed Feb 8 19:08:09 1995
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 236) Date: Tue, 7 Feb 95 22:57 PST
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 237) From: Randolph Bentson <bentson@grieg.seaslug.org>
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 238) To: guru@ncm.com
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 239) Subject: EQL driver tests
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 240)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 241)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 242) I have been checking out your eql driver. (Nice work, that!)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 243) Although you may already done this performance testing, here
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 244) are some data I've discovered.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 245)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 246) Randolph Bentson
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 247) bentson@grieg.seaslug.org
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 248)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 249) ------------------------------------------------------------------
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 250)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 251)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 252) A pseudo-device driver, EQL, written by Simon Janes, can be used
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 253) to bundle multiple SLIP connections into what appears to be a
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 254) single connection. This allows one to improve dial-up network
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 255) connectivity gradually, without having to buy expensive DSU/CSU
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 256) hardware and services.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 257)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 258) I have done some testing of this software, with two goals in
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 259) mind: first, to ensure it actually works as described and
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 260) second, as a method of exercising my device driver.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 261)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 262) The following performance measurements were derived from a set
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 263) of SLIP connections run between two Linux systems (1.1.84) using
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 264) a 486DX2/66 with a Cyclom-8Ys and a 486SLC/40 with a Cyclom-16Y.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 265) (Ports 0,1,2,3 were used. A later configuration will distribute
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 266) port selection across the different Cirrus chips on the boards.)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 267) Once a link was established, I timed a binary ftp transfer of
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 268) 289284 bytes of data. If there were no overhead (packet headers,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 269) inter-character and inter-packet delays, etc.) the transfers
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 270) would take the following times::
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 271)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 272) bits/sec seconds
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 273) 345600 8.3
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 274) 234600 12.3
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 275) 172800 16.7
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 276) 153600 18.8
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 277) 76800 37.6
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 278) 57600 50.2
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 279) 38400 75.3
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 280) 28800 100.4
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 281) 19200 150.6
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 282) 9600 301.3
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 283)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 284) A single line running at the lower speeds and with large packets
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 285) comes to within 2% of this. Performance is limited for the higher
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 286) speeds (as predicted by the Cirrus databook) to an aggregate of
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 287) about 160 kbits/sec. The next round of testing will distribute
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 288) the load across two or more Cirrus chips.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 289)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 290) The good news is that one gets nearly the full advantage of the
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 291) second, third, and fourth line's bandwidth. (The bad news is
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 292) that the connection establishment seemed fragile for the higher
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 293) speeds. Once established, the connection seemed robust enough.)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 294)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 295) ====== ======== === ======== ======= ======= ===
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 296) #lines speed mtu seconds theory actual %of
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 297) kbit/sec duration speed speed max
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 298) ====== ======== === ======== ======= ======= ===
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 299) 3 115200 900 _ 345600
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 300) 3 115200 400 18.1 345600 159825 46
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 301) 2 115200 900 _ 230400
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 302) 2 115200 600 18.1 230400 159825 69
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 303) 2 115200 400 19.3 230400 149888 65
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 304) 4 57600 900 _ 234600
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 305) 4 57600 600 _ 234600
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 306) 4 57600 400 _ 234600
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 307) 3 57600 600 20.9 172800 138413 80
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 308) 3 57600 900 21.2 172800 136455 78
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 309) 3 115200 600 21.7 345600 133311 38
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 310) 3 57600 400 22.5 172800 128571 74
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 311) 4 38400 900 25.2 153600 114795 74
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 312) 4 38400 600 26.4 153600 109577 71
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 313) 4 38400 400 27.3 153600 105965 68
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 314) 2 57600 900 29.1 115200 99410.3 86
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 315) 1 115200 900 30.7 115200 94229.3 81
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 316) 2 57600 600 30.2 115200 95789.4 83
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 317) 3 38400 900 30.3 115200 95473.3 82
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 318) 3 38400 600 31.2 115200 92719.2 80
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 319) 1 115200 600 31.3 115200 92423 80
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 320) 2 57600 400 32.3 115200 89561.6 77
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 321) 1 115200 400 32.8 115200 88196.3 76
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 322) 3 38400 400 33.5 115200 86353.4 74
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 323) 2 38400 900 43.7 76800 66197.7 86
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 324) 2 38400 600 44 76800 65746.4 85
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 325) 2 38400 400 47.2 76800 61289 79
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 326) 4 19200 900 50.8 76800 56945.7 74
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 327) 4 19200 400 53.2 76800 54376.7 70
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 328) 4 19200 600 53.7 76800 53870.4 70
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 329) 1 57600 900 54.6 57600 52982.4 91
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 330) 1 57600 600 56.2 57600 51474 89
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 331) 3 19200 900 60.5 57600 47815.5 83
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 332) 1 57600 400 60.2 57600 48053.8 83
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 333) 3 19200 600 62 57600 46658.7 81
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 334) 3 19200 400 64.7 57600 44711.6 77
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 335) 1 38400 900 79.4 38400 36433.8 94
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 336) 1 38400 600 82.4 38400 35107.3 91
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 337) 2 19200 900 84.4 38400 34275.4 89
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 338) 1 38400 400 86.8 38400 33327.6 86
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 339) 2 19200 600 87.6 38400 33023.3 85
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 340) 2 19200 400 91.2 38400 31719.7 82
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 341) 4 9600 900 94.7 38400 30547.4 79
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 342) 4 9600 400 106 38400 27290.9 71
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 343) 4 9600 600 110 38400 26298.5 68
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 344) 3 9600 900 118 28800 24515.6 85
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 345) 3 9600 600 120 28800 24107 83
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 346) 3 9600 400 131 28800 22082.7 76
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 347) 1 19200 900 155 19200 18663.5 97
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 348) 1 19200 600 161 19200 17968 93
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 349) 1 19200 400 170 19200 17016.7 88
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 350) 2 9600 600 176 19200 16436.6 85
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 351) 2 9600 900 180 19200 16071.3 83
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 352) 2 9600 400 181 19200 15982.5 83
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 353) 1 9600 900 305 9600 9484.72 98
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 354) 1 9600 600 314 9600 9212.87 95
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 355) 1 9600 400 332 9600 8713.37 90
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 356) ====== ======== === ======== ======= ======= ===
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 357)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 358) 5.2. Anthony Healy's Report
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 359) ---------------------------
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 360)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 361) ::
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 362)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 363) Date: Mon, 13 Feb 1995 16:17:29 +1100 (EST)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 364) From: Antony Healey <ahealey@st.nepean.uws.edu.au>
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 365) To: Simon Janes <guru@ncm.com>
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 366) Subject: Re: Load Balancing
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 367)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 368) Hi Simon,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 369) I've installed your patch and it works great. I have trialed
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 370) it over twin SL/IP lines, just over null modems, but I was
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 371) able to data at over 48Kb/s [ISDN link -Simon]. I managed a
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 372) transfer of up to 7.5 Kbyte/s on one go, but averaged around
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 373) 6.4 Kbyte/s, which I think is pretty cool. :)
Orange Pi5 kernel
Deprecated Linux kernel 5.10.110 for OrangePi 5/5B/5+ boards
3 Commits
0 Branches
0 Tags