/[gentoo]/xml/htdocs/doc/en/home-router-howto.xml
Gentoo

Contents of /xml/htdocs/doc/en/home-router-howto.xml

Parent Directory Parent Directory | Revision Log Revision Log


Revision 1.50 - (show annotations) (download) (as text)
Mon Aug 21 10:47:36 2006 UTC (8 years, 11 months ago) by vapier
Branch: MAIN
Changes since 1.49: +15 -5 lines
File MIME type: application/xml
add a small section on ip_dynaddr for Kristian

1 <?xml version='1.0' encoding='UTF-8'?>
2 <!DOCTYPE guide SYSTEM "/dtd/guide.dtd">
3 <!-- $Header: /var/cvsroot/gentoo/xml/htdocs/doc/en/home-router-howto.xml,v 1.49 2006/08/21 10:21:57 vapier Exp $ -->
4
5 <guide link="/doc/en/home-router-howto.xml" lang="en">
6 <title>Home Router Guide</title>
7
8 <author title="Author">
9 <mail link="vapier@gentoo.org">Mike Frysinger</mail>
10 </author>
11
12 <abstract>
13 This document details how to turn an old Gentoo machine into a router
14 for connecting your home network to the internet.
15 </abstract>
16
17 <!-- The content of this document is released into the public domain -->
18 <license/>
19
20 <version>1.31</version>
21 <date>2006-08-21</date>
22
23 <chapter>
24 <title>Introduction</title>
25 <section>
26 <body>
27
28 <p>
29 Building your own router out of old spare parts has many advantages over buying
30 a pre-made canned router by say Linksys. The biggest one by far is control
31 over the connection. The other advantages are left up to your imagination;
32 just about anything can be done in this scenario, it's just a matter of needing
33 it.
34 </p>
35
36 <p>
37 This guide will show you how to setup Network Address Translation (NAT) on the
38 router (kernel and iptables), add and configure common services (Domain Name
39 System (DNS) via dnsmasq, dhcp via dhcpcd, ADSL via rp-pppoe), and conclude
40 with more elaborate and fun things that can be done (port forwarding, traffic
41 shaping, proxies/caching, etc...).
42 </p>
43
44 <p>
45 Before getting started, there's a few basic requirements you must meet. First,
46 you'll need a computer that has at least 2 Network Interface Cards (NICs) in
47 it. Next, you'll need the configuration settings for your internet connection
48 (may include things like IP/DNS/Gateway/username/password). Finally, you'll
49 need a bit of spare time and some Gentoo loving.
50 </p>
51
52 <p>
53 The conventions used in this guide are:
54 </p>
55
56 <ul>
57 <li>eth0 - NIC connected to the Local Area Network (LAN)</li>
58 <li>eth1 - NIC connected to the Wide Area Network (WAN)</li>
59 <li>LAN utilizes the private 192.168.0.xxx network</li>
60 <li>router is hardcoded to the standard 192.168.0.1 IP</li>
61 <li>router is running Linux 2.4 or 2.6; you're on your own with 2.0/2.2</li>
62 </ul>
63
64 <impo>
65 Due to security precautions, I would highly suggest you shut down any unneeded
66 services on the router until we have a chance to get the firewall up and
67 rolling. To view the currently running services, just run <c>rc-status</c>.
68 </impo>
69
70 </body>
71 </section>
72 </chapter>
73
74 <chapter>
75 <title>Kernel setup (know thyself first)</title>
76 <section>
77 <body>
78
79 <p>
80 Your kernel needs to have the drivers running for both your NICs. To see if
81 your cards are already setup, just run <c>ifconfig</c>. Your output may differ
82 slightly from the following, that's fine. What matters is that the interface
83 shows up at all.
84 </p>
85
86 <pre caption="Checking NICs">
87 # <i>ifconfig -a</i>
88 eth0 Link encap:Ethernet HWaddr 00:60:F5:07:07:B8
89 BROADCAST MULTICAST MTU:1500 Metric:1
90 RX packets:0 errors:0 dropped:0 overruns:0 frame:0
91 TX packets:0 errors:0 dropped:0 overruns:0 carrier:0
92 collisions:0 txqueuelen:1000
93 RX bytes:0 (0.0 b) TX bytes:0 (0.0 b)
94 Interrupt:11 Base address:0x9800
95
96 eth1 Link encap:Ethernet HWaddr 00:60:F5:07:07:B9
97 BROADCAST MULTICAST MTU:1500 Metric:1
98 RX packets:0 errors:0 dropped:0 overruns:0 frame:0
99 TX packets:0 errors:0 dropped:0 overruns:0 carrier:0
100 collisions:0 txqueuelen:1000
101 RX bytes:0 (0.0 b) TX bytes:0 (0.0 b)
102 Interrupt:10 Base address:0x9400
103 </pre>
104
105 <p>
106 If you do not see your two cards showing up and you're not sure what kind of
107 cards you have, try running <c>lspci | grep Ethernet</c>. You can get that
108 from <c>emerge pciutils</c>. Once you have this information, go into your
109 kernel and add support for the correct drivers.
110 </p>
111
112 <p>
113 The next thing you'll need is support for iptables and NAT (and packet shaping
114 if you want). The following list is split up into always required (*),
115 required only for adsl via PPPoE (a), suggested for everyone (x), and only
116 for shaper (s) features. It does not matter whether you build the features
117 into the kernel or as a module so long as when the feature is needed, the
118 correct module(s) are loaded (module loading is left to the reader as a fun
119 exercise however).
120 </p>
121
122 <pre caption="Network Options">
123 Networking options ---&gt;
124 [*] TCP/IP networking
125 [*] IP: advanced router
126 [*] Network packet filtering (replaces ipchains)
127 <comment>If you use 2.4.x, you have to enable the following for DHCP:</comment>
128 [*] Socket Filtering
129
130 IP: Netfilter Configuration ---&gt;
131 [*] Connection tracking (required for masq/NAT)
132 [x] FTP protocol support
133 [x] IRC protocol support
134 [*] IP tables support (required for filtering/masq/NAT)
135 [*] IP range match support
136 [x] MAC address match support
137 [*] Multiple port match support
138 [*] Packet filtering
139 [*] REJECT target support
140 [x] REDIRECT target support
141 [*] Full NAT
142 [*] MASQUERADE target support
143 [s] Packet mangling
144 [s] MARK target support
145 [x] LOG target support
146
147 QoS and/or fair queueing ---&gt;
148 [s] QoS and/or fair queueing
149 [s] HTB packet scheduler
150 [s] Ingress Qdisc
151
152 [a] PPP (point-to-point protocol) support
153 [a] PPP filtering
154 [a] PPP support for async serial ports
155 [a] PPP support for sync tty ports
156 [a] PPP Deflate compression
157 [a] PPP BSD-Compress compression
158 [a] PPP over Ethernet
159 </pre>
160
161 <note>
162 Somethings may be slightly different in a 2.4 vs 2.6 kernel, but you should be
163 able to figure it out :).
164 </note>
165
166 </body>
167 </section>
168 </chapter>
169
170 <chapter>
171 <title>Hug the WAN (a.k.a. The Internet)</title>
172
173 <section>
174 <title>Intro</title>
175 <body>
176
177 <p>
178 There are many ways to connect to the internet so I'll just cover the ones I'm
179 familiar with. That leaves us with ADSL (PPPoE) and cable modems
180 (static/dynamic). If there are other methods out there, feel free to write up
181 a little blurb and e-mail me. Feel free to skip any of the following sections
182 in this chapter that don't apply to you. This chapter is just about getting
183 the router connected to the internet via eth1.
184 </p>
185
186 </body>
187 </section>
188 <section>
189 <title>ADSL and PPPoE</title>
190 <body>
191
192 <p>
193 All the fancy PPPoE software has been bundled up into one little nice package
194 nowadays called <uri link="http://www.roaringpenguin.com/">Roaring
195 Penguin</uri>. Simply <c>emerge rp-pppoe</c> and you'll be on your way.
196 Remember how I said you'll need username/password information? Well I wasn't
197 lying so I hope you have it now! Load up <path>/etc/ppp/pppoe.conf</path> in
198 your favorite editor and set it up.
199 </p>
200
201 <note>
202 In order for the following net settings to work, you must have
203 baselayout-1.11.14 or later installed on your system.
204 </note>
205
206 <pre caption="Setting up eth1">
207 <comment>(Replace 'vla9h924' with your username and 'password' with your password)</comment>
208
209 # <i>nano /etc/ppp/pap-secrets</i>
210 <comment># client server secret</comment>
211 "vla9h924" * "password"
212 # <i>nano /etc/conf.d/net</i>
213 <comment>Tell baselayout to use adsl for your eth1:</comment>
214 config_eth1=( "adsl" )
215 user_eth1=( "vla9h924" )
216 # <i>ln -s net.lo /etc/init.d/net.eth1</i>
217 # <i>rc-update add net.eth1 default</i>
218 # <i>/etc/init.d/net.eth1 start</i>
219 </pre>
220
221 <warn>
222 When the DSL interface comes up, it will create ppp0. Although your NIC is
223 called eth1, the IP is actually bound to ppp0. From now on, when you see
224 examples that utilize 'eth1', substitute with 'ppp0'.
225 </warn>
226
227 </body>
228 </section>
229
230 <section>
231 <title>Cable and/or dynamic/static IP</title>
232 <body>
233
234 <p>
235 If you have a static IP then you will need a few more details than if
236 you have a dynamic IP. For static users, you will need your IP,
237 gateway, and DNS servers.
238 </p>
239
240 <pre caption="Setting up eth1">
241 <comment>Dynamic IP Users:</comment>
242 # <i>emerge dhcpcd</i>
243 # <i>nano /etc/conf.d/net</i>
244 <comment>You'll need an entry like so:</comment>
245 config_eth1=( "dhcp" )
246
247 <comment>Static IP Users:</comment>
248 # <i>nano /etc/conf.d/net</i>
249 <comment>You'll need entries like so:</comment>
250 config_eth1=( "66.92.78.102 broadcast 66.92.78.255 netmask 255.255.255.0" )
251 routes_eth1=( "default gw 66.92.78.1" )
252 # <i>nano /etc/resolv.conf</i>
253 <comment>Add one line per DNS server:</comment>
254 nameserver 123.123.123.123
255
256 <comment>Dynamic and Static Setup:</comment>
257 # <i>ln -s net.lo /etc/init.d/net.eth1</i>
258 # <i>rc-update add net.eth1 default</i>
259 # <i>/etc/init.d/net.eth1 start</i>
260 </pre>
261
262 <p>
263 You should be all set to go now.
264 </p>
265
266 </body>
267 </section>
268 </chapter>
269
270 <chapter>
271 <title>Hug the LAN (bring along some friends)</title>
272 <section>
273 <body>
274
275 <p>
276 This step is a breeze compared to the previous one.
277 </p>
278
279 <pre caption="Setting up eth0">
280 # <i>nano /etc/conf.d/net</i>
281 <comment>Add a line like the following:</comment>
282 config_eth0=( "192.168.0.1 broadcast 192.168.0.255 netmask 255.255.255.0" )
283 # <i>rc-update add net.eth0 default</i>
284 # <i>/etc/init.d/net.eth0 start</i>
285 </pre>
286
287 </body>
288 </section>
289 </chapter>
290
291 <chapter>
292 <title>LAN Services (because we're nice people)</title>
293
294 <section>
295 <title>DHCP Server</title>
296 <body>
297
298 <p>
299 I bet it'd be nice if everyone else in your house could just plug their
300 computers into the network and things would just work. No need to remember
301 mind-numbing details or make them stare at confusing configuration screens!
302 Life would be grand eh? Introducing the Dynamic Host Configuration Protocol
303 (DHCP) and why you should care.
304 </p>
305
306 <p>
307 DHCP is exactly what its name implies. It's a protocol that allows you
308 to dynamically configure other hosts automatically. You run a DHCP server on
309 the router, give it all the information about your network (valid IPs,
310 DNS servers, gateways, etc...), and then when the other hosts start up, they
311 run a DHCP client to automatically configure themselves. No fuss, no muss!
312 For more information about DHCP, you can always visit <uri
313 link="http://en.wikipedia.org/wiki/DHCP">Wikipedia</uri>.
314 </p>
315
316 <p>
317 We'll use a package called dnsmasq which provides both DHCP and DNS services.
318 For now lets just focus on the DHCP aspect. Note that if you want to run a
319 different DHCP server, you can find another example in the Fun Things chapter.
320 Also, if you wish to tinker with the DHCP server settings, just read the
321 comments in <path>/etc/dnsmasq.conf</path>. All the defaults should work fine
322 though.
323 </p>
324
325 <pre caption="Setting up a DHCP server">
326 # <i>emerge dnsmasq</i>
327 # <i>nano /etc/dnsmasq.conf</i>
328 <comment>Add this line to enable dhcp:</comment>
329 dhcp-range=192.168.0.100,192.168.0.250,72h
330 <comment>Restrict dnsmasq to just the LAN interface</comment>
331 interface=eth0
332
333 # <i>rc-update add dnsmasq default</i>
334 # <i>/etc/init.d/dnsmasq start</i>
335 </pre>
336
337 <p>
338 Now your little router is a bona-fide DHCP server! Plugin those computers and
339 watch them work! With Windows systems you should go into the TCP/IP Properties
340 and select the 'Obtain an IP address automatically' and 'Obtain DNS server
341 address automatically' options. Sometimes the changes aren't instantaneous, so
342 you may have to open a command prompt and run <c>ipconfig /release</c> and
343 <c>ipconfig /renew</c>. But enough about Windows, let's get back to our
344 favorite penguin.
345 </p>
346
347 </body>
348 </section>
349
350 <section>
351 <title>DNS Server</title>
352 <body>
353
354 <p>
355 When people want to visit a place on the internet, they remember names, not a
356 string of funky numbers. After all, what's easier to remember, ebay.com or
357 66.135.192.87? This is where the DNS steps in. DNS servers run all over the
358 internet, and whenever someone wants to visit 'ebay.com', these servers turn
359 'ebay.com' (what we understand) into '66.135.192.87' (what our computers
360 understand). For more information about DNS, you can always visit <uri
361 link="http://en.wikipedia.org/wiki/DNS">Wikipedia</uri>.
362 </p>
363
364 <p>
365 Since we're using dnsmasq for our DHCP server, and it includes a DNS server,
366 you've got nothing left to do here! Your little router is already providing
367 DNS to its DHCP clients. Bet you wish everything was this easy ;).
368 </p>
369
370 <p>
371 You're welcome to choose other DNS servers if you're more comfortable with
372 them, but the reason dnsmasq is great is because it was designed to do exactly
373 what we want and nothing more. It's a little DNS caching/forwarding server for
374 local networks. We're not looking to provide DNS for our own domain here, just
375 offer simple DNS services to everyone else on our LAN.
376 </p>
377
378 </body>
379 </section>
380
381 <section>
382 <title>NAT (a.k.a. IP-masquerading)</title>
383 <body>
384
385 <p>
386 At this point, people on your network can talk to each other and they can look
387 up hostnames via DNS, but they still can't actually connect to the internet.
388 While you may think that's great (more bandwidth for you!), I bet they're not
389 too happy just yet.
390 </p>
391
392 <p>
393 This is where Network Address Translation (NAT) steps in. NAT is a way of
394 connecting multiple computers in a private LAN to the internet when you have a
395 smaller number of public IP addresses available to you. Typically you are given
396 1 IP by your ISP, but you want to let your whole house connect to the internet.
397 NAT is the magic that makes this possible. For more information about NAT, you
398 can always visit <uri link="http://en.wikipedia.org/wiki/NAT">Wikipedia</uri>.
399 </p>
400
401 <note>
402 Before we get started, make sure you have iptables on your system. Although it
403 is automatically installed on most systems, you may not have it. If you don't,
404 just run <c>emerge iptables</c>.
405 </note>
406
407 <pre caption="Setting up iptables">
408 <comment>First we flush our current rules</comment>
409 # <i>iptables -F</i>
410 # <i>iptables -t nat -F</i>
411
412 <comment>Setup default policies to handle unmatched traffic</comment>
413 # <i>iptables -P INPUT ACCEPT</i>
414 # <i>iptables -P OUTPUT ACCEPT</i>
415 # <i>iptables -P FORWARD DROP</i>
416
417 <comment>Copy and paste these examples ...</comment>
418 # <i>export LAN=eth0</i>
419 # <i>export WAN=eth1</i>
420
421 <comment>Then we lock our services so they only work from the LAN</comment>
422 # <i>iptables -I INPUT 1 -i ${LAN} -j ACCEPT</i>
423 # <i>iptables -I INPUT 1 -i lo -j ACCEPT</i>
424 # <i>iptables -A INPUT -p UDP --dport bootps -i ! ${LAN} -j REJECT</i>
425 # <i>iptables -A INPUT -p UDP --dport domain -i ! ${LAN} -j REJECT</i>
426
427 <comment>(Optional) Allow access to our ssh server from the WAN</comment>
428 # <i>iptables -A INPUT -p TCP --dport ssh -i ${WAN} -j ACCEPT</i>
429
430 <comment>Drop TCP / UDP packets to privileged ports</comment>
431 # <i>iptables -A INPUT -p TCP -i ! ${LAN} -d 0/0 --dport 0:1023 -j DROP</i>
432 # <i>iptables -A INPUT -p UDP -i ! ${LAN} -d 0/0 --dport 0:1023 -j DROP</i>
433
434 <comment>Finally we add the rules for NAT</comment>
435 # <i>iptables -I FORWARD -i ${LAN} -d 192.168.0.0/255.255.0.0 -j DROP</i>
436 # <i>iptables -A FORWARD -i ${LAN} -s 192.168.0.0/255.255.0.0 -j ACCEPT</i>
437 # <i>iptables -A FORWARD -i ${WAN} -d 192.168.0.0/255.255.0.0 -j ACCEPT</i>
438 # <i>iptables -t nat -A POSTROUTING -o ${WAN} -j MASQUERADE</i>
439 <comment>Tell the kernel that ip forwarding is OK</comment>
440 # <i>echo 1 > /proc/sys/net/ipv4/ip_forward</i>
441 # <i>for f in /proc/sys/net/ipv4/conf/*/rp_filter ; do echo 1 > $f ; done</i>
442
443 <comment>This is so when we boot we don't have to run the rules by hand</comment>
444 # <i>/etc/init.d/iptables save</i>
445 # <i>rc-update add iptables default</i>
446 # <i>nano /etc/sysctl.conf</i>
447 <comment>Add/Uncomment the following lines:</comment>
448 net.ipv4.ip_forward = 1
449 net.ipv4.conf.default.rp_filter = 1
450
451 <comment>If you have a dynamic internet address you probably want to enable this:</comment>
452 net.ipv4.ip_dynaddr = 1
453 </pre>
454
455 <p>
456 Once you've typed out all of that, the rest of your network should now be able
457 to use the internet as if they were directly connected themselves.
458 </p>
459
460 <p>
461 The ip_dynaddr option is useful for dial on demand systems or when your ISP
462 gives out dynamic addresses. This works around the problem where a connection
463 is attempted before the internet interface is fully setup. Really this just
464 provides for a smoother network experience for users behind your router.
465 </p>
466
467 </body>
468 </section>
469 </chapter>
470
471 <chapter>
472 <title>Fun Things (for a rainy day)</title>
473
474 <section>
475 <title>Intro</title>
476 <body>
477
478 <p>
479 Believe it or not, you're done :). From here on out, I'll cover a bunch of
480 common topics that may interest you. Everything in this chapter is completely
481 optional.
482 </p>
483
484 </body>
485 </section>
486
487 <section>
488 <title>Port Forwarding</title>
489 <body>
490
491 <p>
492 Sometimes you would like to be able to host services on a computer behind the
493 router, or just to make your life easier when connecting remotely. Perhaps you
494 want to run a FTP, HTTP, SSH, or VNC server on one or more machines behind your
495 router and be able to connect to them all. The only caveat is that you can
496 only have one service/machine combo per port. For example, there is no
497 practical way to setup three FTP servers behind your router and then try to
498 connect to them all through port 21; only one can be on port 21 while the
499 others would have to be on say port 123 and port 567.
500 </p>
501
502 <p>
503 All the port forwarding rules are of the form <c>iptables -t nat -A PREROUTING
504 [-p protocol] --dport [external port on router] -i ${WAN} -j DNAT --to [ip/port
505 to forward to]</c>. Unfortunately, iptables does not accept hostnames when port
506 forwarding. If you are forwarding an external port to the same port on the
507 internal machine, you can omit the destination port. See the iptables(8) man
508 page for more information.
509 </p>
510
511 <pre caption="Running the iptables commands">
512 <comment>Copy and paste these examples ...</comment>
513 # <i>export LAN=eth0</i>
514 # <i>export WAN=eth1</i>
515
516 <comment>Forward port 2 to ssh on an internal host</comment>
517 # <i>iptables -t nat -A PREROUTING -p tcp --dport 2 -i ${WAN} -j DNAT --to 192.168.0.2:22</i>
518
519 <comment>FTP forwarding to an internal host</comment>
520 # <i>iptables -t nat -A PREROUTING -p tcp --dport 21 -i ${WAN} -j DNAT --to 192.168.0.56</i>
521
522 <comment>HTTP forwarding to an internal host</comment>
523 # <i>iptables -t nat -A PREROUTING -p tcp --dport 80 -i ${WAN} -j DNAT --to 192.168.0.56</i>
524
525 <comment>VNC forwarding for internal hosts</comment>
526 # <i>iptables -t nat -I PREROUTING -p tcp --dport 5900 -i ${WAN} -j DNAT --to 192.168.0.2</i>
527 # <i>iptables -t nat -I PREROUTING -p tcp --dport 5901 -i ${WAN} -j DNAT --to 192.168.0.3:5900</i>
528 <comment>If you want to VNC in to 192.168.0.3, then just add ':1' to the router's hostname</comment>
529
530 <comment>Bittorrent forwarding</comment>
531 # <i>iptables -t nat -A PREROUTING -p tcp --dport 6881:6889 -i ${WAN} -j DNAT --to 192.168.0.2</i>
532
533 <comment>eDonkey/eMule forwarding</comment>
534 # <i>iptables -t nat -A PREROUTING -p tcp --dport 4662 -i ${WAN} -j DNAT --to 192.168.0.55</i>
535
536 <comment>Game Cube Warp Pipe support</comment>
537 # <i>iptables -t nat -A PREROUTING -p udp --dport 4000 -i ${WAN} -j DNAT --to 192.168.0.56</i>
538
539 <comment>Playstation 2 Online support</comment>
540 # <i>iptables -t nat -A PREROUTING -p tcp --dport 10070:10080 -i ${WAN} -j DNAT --to 192.168.0.11</i>
541 # <i>iptables -t nat -A PREROUTING -p udp --dport 10070:10080 -i ${WAN} -j DNAT --to 192.168.0.11</i>
542
543 <comment>Xbox Live</comment>
544 # <i>iptables -t nat -A PREROUTING -p tcp --dport 3074 -i ${WAN} -j DNAT --to 192.168.0.69</i>
545 # <i>iptables -t nat -A PREROUTING -p udp --dport 3074 -i ${WAN} -j DNAT --to 192.168.0.69</i>
546 # <i>iptables -t nat -A PREROUTING -p udp --dport 88 -i ${WAN} -j DNAT --to 192.168.0.69</i>
547 </pre>
548
549 <note>
550 If you have other common / cool examples, please <mail
551 link="vapier@gentoo.org">e-mail me</mail>.
552 </note>
553
554 </body>
555 </section>
556
557 <section>
558 <title>Identd (for IRC)</title>
559 <body>
560
561 <p>
562 Internet Relay Chat utilizes the ident service pretty heavily. Now that the
563 IRC clients are behind the router, we need a way to host ident for both the
564 router and the clients. One such server has been created called
565 <c>midentd</c>.
566 </p>
567
568 <pre caption="Setting up ident">
569 # <i>emerge midentd</i>
570 # <i>rc-update add midentd default</i>
571 # <i>/etc/init.d/midentd start</i>
572 </pre>
573
574 <p>
575 There are a few other ident servers in portage. Depending on your needs, I
576 would recommend checking out <c>oidentd</c> and <c>fakeidentd</c>.
577 </p>
578
579 </body>
580 </section>
581
582 <!--
583 <section>
584 <title>Traffic Shaping</title>
585 <body>
586 <p>
587 This is an attempt to simply and Gentooify the <uri link="http://www.tldp.org/HOWTO/ADSL-Bandwidth-Management-HOWTO/">ADSL Bandwidth Management HOWTO</uri>
588 found over at the TLDP. Feel free to refer to the original document
589 for more details.
590 </p>
591
592 <p>
593 Here we will be setting up what some people refer to as a "Packet Shaper",
594 <uri link="http://en.wikipedia.org/wiki/Traffic_shaping">"Traffic Shaping"</uri>,
595 or <uri link="http://en.wikipedia.org/wiki/QoS">"Quality of Service"</uri>.
596 Simply put, we want to setup rules on our router that will slow down
597 certain activities (like sending large e-mails or downloading from P2P
598 networks) while keeping other activities (like browsing the web or playing
599 online video games) reasonably fast. A 30 second difference in a video
600 game is a lot worse than a 30 second difference in downloading large
601 files :).
602 </p>
603
604 <p>
605 The first thing is to make sure your kernel has all the features added to
606 it. See the chapter on <uri link="#doc_chap2">Kernel setup</uri> for more
607 information. Next, you will need to <c>emerge iptables iputils</c> so that
608 you will have access to the <c>iptables</c>, <c>ip</c>, and <c>tc</c>
609 commands.
610 </p>
611
612 <p>
613 Before we jump into the commands, let's cover a little of the theory. The
614 way this whole system works is to classify common network streams and then
615 to prioritize them. You use iptables to classify network streams, iputils
616 to define the different priority levels, and the kernel to adjust speeds.
617 Just remember that although you can control outbound traffic pretty tightly
618 (from the LAN to the WAN), your ability to control inbound traffic (from
619 the WAN to the LAN) is somewhat limited. Just remember that the following
620 examples are to get your feet wet; if you want more then I'd suggest
621 reading up on the subject. In this example, we will be using the
622 <uri link="http://luxik.cdi.cz/~devik/qos/htb/">Hierarchical Token Buckets (HTB)</uri>
623 packet scheduling algorithm. Still with me? Great, let's start shaping :).
624 </p>
625
626 <pre caption="Setup">
627 DEV=eth1 <comment>NIC connected to WAN</comment>
628 RATE_OUT=100 <comment>Available outbound bandwidth (in kilobits [kb])</comment>
629 RATE_IN=1400 <comment>Available inbound bandwidth (in kb)</comment>
630
631 <comment>Here we initialize the priority system. The 45 is used to set the default classification level.</comment>
632 ip link set dev ${DEV} qlen 30
633 tc qdisc add dev ${DEV} root handle 1: htb default 45
634 tc class add dev ${DEV} parent 1: classid 1:1 htb rate ${RATE_OUT}kbit
635 </pre>
636
637 <p>
638 Here we initialized the system which will be used to prioritize all of
639 our network traffic. We created our queue, told it to use the HTB
640 algorithm, and set the default classification level to '45'. The
641 default is completely arbitrary, as are the levels we choose from
642 here on out. The only thing that matters is how the levels compare
643 relatively; a level '10' packet will be given preference over a
644 level '45' packet. Let's move on to declaring different levels.
645 </p>
646
647 <pre caption="Declaring levels">
648 tc class add dev $DEV parent 1:1 classid 1:10 htb rate $rkbit ceil $tkbit prio $p
649 tc qdisc add dev $DEV parent 1:10 handle 10: sfq
650 </pre>
651 </body>
652 </section>
653 -->
654
655 <section>
656 <title>Time Server</title>
657 <body>
658
659 <p>
660 Keeping your system time correct is essential in maintaining a healthy system.
661 One of the most common ways of accomplishing this is with the Network Time
662 Protocol (NTP) and the ntp package (which provides implementations for both
663 server and client).
664 </p>
665
666 <p>
667 Many people run ntp clients on their computers. Obviously, the more clients in
668 the world, the larger the load the ntp servers need to shoulder. In
669 environments like home networks though, we can help keep the load down on
670 public servers while still providing the proper time to all our computers. As
671 an added bonus, our private updates will be a lot faster for the clients too!
672 All we have to do is run a ntp server on our router that synchronizes itself
673 with the public internet servers while providing the time to the rest of the
674 computers in the network. To get started, simply <c>emerge ntp</c> on the
675 router.
676 </p>
677
678 <pre caption="Setting up the NTP server">
679 # <i>nano /etc/conf.d/ntp-client</i>
680 <comment>Customize if you wish but the defaults should be fine</comment>
681 # <i>rc-update add ntp-client default</i>
682
683 # <i>nano /etc/ntp.conf</i>
684 <comment>Add the follwing lines:</comment>
685 restrict default ignore
686 restrict 192.168.0.0 mask 255.255.255.0 notrust nomodify notrap
687 <comment>These will allow only ntp clients with an IP
688 address in the 192.168.0.xxx range to use your ntp server</comment>
689 # <i>nano /etc/conf.d/ntpd</i>
690 <comment>Customize if you wish but the defaults should be fine</comment>
691 # <i>rc-update add ntpd default</i>
692
693 # <i>/etc/init.d/ntp-client start</i>
694 # <i>/etc/init.d/ntpd start</i>
695 </pre>
696
697 <note>
698 You should make sure that you allow inbound and outbound communication on the
699 ntp port (123/udp) when setting up the server. The client just needs outbound
700 access on port 123 over udp.
701 </note>
702
703 <p>
704 Now, on your clients, have them <c>emerge ntp</c> also. However, we will just
705 run the ntp client so setup is a lot simpler.
706 </p>
707
708 <pre caption="Setting up a NTP client">
709 # <i>nano /etc/conf.d/ntp-client</i>
710 <comment>Change the 'pool.ntp.org' server in the NTPCLIENT_OPTS variable to '192.168.0.1'</comment>
711 # <i>rc-update add ntp-client default</i>
712 # <i>/etc/init.d/ntp-client start</i>
713 </pre>
714
715 </body>
716 </section>
717
718 <section>
719 <title>Rsync Server</title>
720 <body>
721
722 <p>
723 For those who run multiple Gentoo boxes on the same lan, you often want to
724 keep from having every machine running <c>emerge sync</c> with remote
725 servers. By setting up a local rsync, you save on both your bandwidth and
726 the Gentoo rsync servers' bandwidth. It's pretty simple to do.
727 </p>
728
729 <note>
730 For a much more in-depth rsync guide, please see the official <uri
731 link="/doc/en/rsync.xml#local">rsync guide</uri>.
732 </note>
733
734 <p>
735 Since every Gentoo machine requires rsync, theres no need to emerge it. Edit
736 the default <path>/etc/rsyncd.conf</path> config file, uncomment the
737 <c>[gentoo-portage]</c> section, and make sure you add an <c>address</c>
738 option. All the other defaults should be fine.
739 </p>
740
741 <pre caption="Rsync server config">
742 pid file = /var/run/rsyncd.pid
743 use chroot = yes
744 read only = yes
745 address = 192.168.0.1
746
747 [gentoo-portage]
748 path = /mnt/space/portage
749 comment = Gentoo Linux Portage tree
750 exclude = /distfiles /packages
751 </pre>
752
753 <p>
754 Then you need to start the service (again, the defaults are OK).
755 </p>
756
757 <pre caption="Starting the rsync server">
758 # <i>/etc/init.d/rsyncd start</i>
759 # <i>rc-update add rsyncd default</i>
760 </pre>
761
762 <p>
763 Only thing left is to set tell your clients to sync against the router.
764 </p>
765
766 <pre caption="Client SYNC settings in make.conf">
767 SYNC="rsync://192.168.0.1/gentoo-portage"
768 </pre>
769
770 </body>
771 </section>
772
773 <section>
774 <title>Mail Server</title>
775 <body>
776
777 <p>
778 Sometimes it's nice to run your own Simple Mail Transfer Protocol (SMTP) server
779 on the router. You may have your own reason for wanting to do so, but I run it
780 so that the users see mail as being sent instantly and the work of
781 retrying/routing is left up to the mail server. Some ISPs also don't allow for
782 mail relaying for accounts that aren't part of their network (like Verizon).
783 Also, you can easily throttle the delivery of mail so that large attachments
784 won't seriously lag your connection for half an hour.
785 </p>
786
787 <pre caption="Setting up SMTP">
788 # <i>emerge qmail</i>
789 <comment>make sure the output of `hostname` is correct</comment>
790 # <i>ebuild /var/db/pkg/*-*/qmail-1.03-r*/*.ebuild config</i>
791 # <i>iptables -I INPUT -p tcp --dport smtp -i ! ${LAN} -j REJECT</i>
792 # <i>ln -s /var/qmail/supervise/qmail-send /service/qmail-send</i>
793 # <i>ln -s /var/qmail/supervise/qmail-smtpd /service/qmail-smtpd</i>
794 <!--
795 # <i>cd /etc/tcprules.d</i>
796 # <i>nano tcp.qmail-smtp</i>
797 -->
798 # <i>cd /etc</i>
799 # <i>nano tcp.smtp</i>
800 <comment>Add an entry like so to the allow section:</comment>
801 192.168.0.:allow,RELAYCLIENT=""
802 <!--
803 # <i>tcprules tcp.qmail-qmtp.cdb rules.tmp &lt; tcp.qmail-smtp</i>
804 -->
805 # <i>tcprules tcp.smtp.cdb rules.tmp &lt; tcp.smtp</i>
806 # <i>rc-update add svscan default</i>
807 # <i>/etc/init.d/svscan start</i>
808 </pre>
809
810 <p>
811 I'm a huge fan of qmail, but you're free to use a different mta :). When you
812 setup e-mail on the hosts in your network, tell them that their SMTP server is
813 192.168.0.1 and everything should be peachy. You might want to visit the <uri
814 link="http://qmail.org/">qmail homepage</uri> for more documentation.
815 </p>
816
817 </body>
818 </section>
819
820 <!--
821 <section>
822 <title>E-mail Virus Scanning</title>
823 <body>
824 <p>
825 If you'd like to provide e-mail virus scanning for your users, but
826 don't want to have to install a virus scanner on every single machine,
827 then <c>pop3vscan</c> may just be the thing for you; a transparent
828 Post Office Protocol (POP) scanner.
829 </p>
830
831 <pre caption="Setting up pop3vscan">
832 TODO
833 </pre>
834
835 </body>
836 </section>
837 -->
838
839 <section>
840 <title>Full DHCP Server</title>
841 <body>
842
843 <p>
844 Earlier we used dnsmasq to provide DHCP service to all our clients. For most
845 people with a simple small LAN, this is perfect. But you may need something
846 with more features. Thus we turn to a full-featured DHCP server as provided
847 by the <uri link="http://www.isc.org/products/DHCP">ISC</uri> folks.
848 </p>
849
850 <pre caption="Setting up dhcpd">
851 # <i>emerge dhcp</i>
852 # <i>nano /etc/dhcp/dhcpd.conf</i>
853 <comment>(Here is a sample configuration file:)</comment>
854 authoritative;
855 ddns-update-style interim;
856 subnet 192.168.0.0 netmask 255.255.255.0 {
857 range 192.168.0.100 192.168.0.250;
858 default-lease-time 259200;
859 max-lease-time 518400;
860 option subnet-mask 255.255.255.0;
861 option broadcast-address 192.168.0.255;
862 option routers 192.168.0.1;
863 option domain-name-servers 192.168.0.1;
864 }
865 # <i>nano /etc/conf.d/dhcpd</i>
866 <comment>(Set IFACE="eth0")</comment>
867 # <i>rc-update add dhcpd default</i>
868 # <i>/etc/init.d/dhcpd start</i>
869 </pre>
870
871 <p>
872 This is the minimal setup required to replace the dnsmasq DHCP functionality
873 that we used earlier. Speaking of which, you did remember to disable the DHCP
874 features in dnsmasq didn't you? If not, you should do so now (just comment
875 out the <c>dhcp-range</c> setting in <path>/etc/dnsmasq.conf</path> and restart
876 the service).
877 </p>
878
879 </body>
880 </section>
881
882 <section>
883 <title>Connect Another LAN (or two or three or ...)</title>
884 <body>
885
886 <p>
887 Sometimes you have need of connecting the router to another LAN. Maybe you
888 want to hook up a group of friends temporarily, or you're a neat freak and
889 want to section off different groups of computers, or you're just really
890 really bored. Whatever the reasons, extending the router to other LAN
891 networks should be pretty straightforward. In the following examples, I will
892 assume that this new network is connected via a third ethernet card, namely
893 <c>eth2</c>.
894 </p>
895
896 <p>
897 First you need to configure the interface. Just take the instructions in the
898 <uri link="#doc_chap4_pre1">4.1 code listing</uri> and replace <c>eth0</c>
899 with <c>eth2</c> and <c>192.168.0</c> with <c>192.168.1</c>.
900 </p>
901
902 <p>
903 Then you need to tweak dnsmasq to service the new interface. Just edit the
904 <path>/etc/conf.d/dnsmasq</path> file again and append <c>-i eth2</c> to
905 DNSMASQ_OPTS; using -i multiple times is OK. Then edit
906 <path>/etc/dnsmasq.conf</path> and add another line like the dhcp-range line
907 in the <uri link="#doc_chap5_pre1">5.1 code listing</uri>, replacing
908 <c>192.168.0</c> with <c>192.168.1</c>. Having multiple dhcp-range lines is
909 OK too.
910 </p>
911
912 <p>
913 Finally, see the rules in the <uri link="#doc_chap5_pre2">5.2 code
914 listing</uri> and duplicate the rules that have <c>-i ${LAN}</c> in them. You
915 may want to create another variable, say <c>LAN2</c>, to make things easier.
916 </p>
917
918 </body>
919 </section>
920
921 </chapter>
922
923 <chapter>
924 <title>Troubleshooting</title>
925
926 <section>
927 <title>Useful Tools</title>
928 <body>
929
930 <p>
931 If you're having trouble getting your computers to communicate, you may way to
932 try out the following tools (they can all be found in the <c>net-analyzer</c>
933 portage category):
934 </p>
935
936 <table>
937 <tr>
938 <th>Utility</th>
939 <th>Description</th>
940 </tr>
941 <tr>
942 <ti>ethereal</ti>
943 <ti>GUI tool to view all raw network data according to filters</ti>
944 </tr>
945 <tr>
946 <ti>tcpdump</ti>
947 <ti>Console tool to dump all raw network data according to filters</ti>
948 </tr>
949 <tr>
950 <ti>iptraf</ti>
951 <ti>ncurses based IP LAN monitor</ti>
952 </tr>
953 <tr>
954 <ti>ettercap</ti>
955 <ti>ncurses based network monitor/control</ti>
956 </tr>
957 </table>
958
959 </body>
960 </section>
961
962 <section>
963 <title>DHCP Fails To Start</title>
964 <body>
965
966 <p>
967 When starting the dhcp init.d script for the first time, it may fail to load
968 but neglect to give you any useful info.
969 </p>
970
971 <pre caption="DHCP Failing Example">
972 # <i>/etc/init.d/dhcp start</i>
973 * Setting ownership on dhcp.leases ... [ ok ]
974 * Starting dhcpd ... [ !! ]
975 </pre>
976
977 <p>
978 The trick is to know where dhcpd is sending its output. Simply browse to
979 <path>/var/log</path> and read the log files. Since the exact log file depends
980 on the package you are using as a syslog, try running <c>grep -Rl dhcpd
981 /var/log</c> to narrow down the possibilities. Chances are you made a typo in
982 your config file. You could also try running <c>dhcpd -d -f</c> (short for
983 debug / foreground) and debug the error based upon the output.
984 </p>
985
986 </body>
987 </section>
988
989 <section>
990 <title>Incorrect MTU Value</title>
991 <body>
992
993 <p>
994 If you experience odd errors (such as not being some webpages while others
995 load fine), you may be having Path MTU Discovery trouble. The quick way to
996 test is to run this iptables command:
997 </p>
998
999 <pre caption="Circumvent MTU issues">
1000 # <i>iptables -A FORWARD -p tcp --tcp-flags SYN,RST SYN -j TCPMSS --clamp-mss-to-pmtu</i>
1001 </pre>
1002
1003 <p>
1004 This will affect all new connections, so just refresh the website you're
1005 having problems with in order to test. In case it helps, the standard MTU
1006 value for 100mbit ethernet connections is <c>1500</c> while for PPPoE
1007 connections it is <c>1492</c>. For more info, you should read Chapter 15
1008 of the <uri link="http://lartc.org/howto/">Linux Advanced Routing &amp;
1009 Traffic Control HOWTO</uri>.
1010 </p>
1011
1012 </body>
1013 </section>
1014
1015 <section>
1016 <title>Unable to connect two machines directly</title>
1017 <body>
1018
1019 <p>
1020 If (for whatever reason) you want to connect two machines directly together
1021 without a hub or switch, a regular ethernet cable will likely not work, unless
1022 you have an Auto MDI/MDI-X (also known as "autosensing") capable network
1023 adapter. You will need a different cable called a crossover cable. This <uri
1024 link="http://en.wikipedia.org/wiki/Ethernet_crossover_cable">Wikipedia</uri>
1025 page explains the low level details.
1026 </p>
1027
1028 </body>
1029 </section>
1030
1031 </chapter>
1032
1033 <chapter>
1034 <title>Final Notes</title>
1035 <section>
1036 <body>
1037
1038 <p>
1039 I have no final notes other than if you experience any troubles with the guide,
1040 please contact <mail link="vapier@gentoo.org">me</mail> or file a bug with <uri
1041 link="http://bugs.gentoo.org/">Gentoo's Bugtracking Website</uri>. If you have
1042 some interesting bits you think would enhance this guide, by all means send it
1043 my way for inclusion.
1044 </p>
1045
1046 </body>
1047 </section>
1048 </chapter>
1049 </guide>

  ViewVC Help
Powered by ViewVC 1.1.20