Debian Codegraph @

man/teamd.conf.5

raw
1 .TH TEAMD.CONF 5 "2013-07-09" "libteam" "Team daemon configuration"
2 .SH NAME
3 teamd.conf \(em libteam daemon configuration file
4 .SH DESCRIPTION
5 .PP
6 teamd uses JSON format configuration.
7 .SH OPTIONS
8 .TP
9 .BR "device " (string)
10 Desired name of new team device.
11 .TP
12 .BR "debug_level " (int)
13 Level of debug messages. The higher it is the more debug messages will be printed. It is the same as adding "-g" command line options.
14 .RS 7
15 .PP
16 Default:
17 .BR "0"
18 (disabled)
19 .RE
20 .TP
21 .BR "hwaddr " (string)
22 Desired hardware address of new team device. Usual MAC address format is accepted.
23 .TP
24 .BR "runner.name " (string)
25 Name of team device. The following runners are available:
26 .RS 7
27 .PP
28 .BR "broadcast "\(em
29 Simple runner which directs the team device to transmit packets via all ports.
30 .PP
31 .BR "roundrobin "\(em
32 Simple runner which directs the team device to transmits packets in a round-robin fashion.
33 .PP
34 .BR "random "\(em
35 Simple runner which directs the team device to transmits packets on a randomly selected port.
36 .PP
37 .BR "activebackup "\(em
38 Watches for link changes and selects active port to be used for data transfers.
39 .PP
40 .BR "loadbalance "\(em
41 To do passive load balancing, runner only sets up BPF hash function which will determine port for packet transmit. To do active load balancing, runner moves hashes among available ports trying to reach perfect balance.
42 .PP
43 .BR "lacp "\(em
44 Implements 802.3ad LACP protocol. Can use same Tx port selection possibilities as loadbalance runner.
45 .PP
46 Default:
47 .BR "roundrobin"
48 .RE
49 .TP
50 .BR "notify_peers.count " (int)
51 Number of bursts of unsolicited NAs and gratuitous ARP packets sent after port is enabled or disabled.
52 .RS 7
53 .PP
54 Default:
55 .BR "0"
56 (disabled)
57 .PP
58 Default for activebackup runner:
59 .BR "1"
60 .RE
61 .TP
62 .BR "notify_peers.interval " (int)
63 Value is positive number in milliseconds. Specifies an interval between bursts of notify-peer packets.
64 .RS 7
65 .PP
66 Default:
67 .BR "0"
68 .RE
69 .TP
70 .BR "mcast_rejoin.count " (int)
71 Number of bursts of multicast group rejoin requests sent after port is enabled or disabled.
72 .RS 7
73 .PP
74 Default:
75 .BR "0"
76 (disabled)
77 .PP
78 Default for activebackup runner:
79 .BR "1"
80 .RE
81 .TP
82 .BR "mcast_rejoin.interval " (int)
83 Value is positive number in milliseconds. Specifies an interval between bursts of multicast group rejoin requests.
84 .RS 7
85 .PP
86 Default:
87 .BR "0"
88 .RE
89 .TP
90 .BR "link_watch.name "| " ports.PORTIFNAME.link_watch.name " (string)
91 Name of link watcher to be used. The following link watchers are available:
92 .RS 7
93 .PP
94 .BR "ethtool "\(em
95 Uses Libteam lib to get port ethtool state changes.
96 .PP
97 .BR "arp_ping "\(em
98 ARP requests are sent through a port. If an ARP reply is received, the link is considered to be up.
99 .PP
100 .BR "nsna_ping "\(em
101 Similar to the previous, except that it uses IPv6 Neighbor Solicitation / Neighbor Advertisement mechanism. This is an alternative to arp_ping and becomes handy in pure-IPv6 environments.
102 .RE
103 .TP
104 .BR "ports " (object)
105 List of ports, network devices, to be used in a team device.
106 .RS 7
107 .PP
108 See examples for more information.
109 .RE
110 .TP
111 .BR "ports.PORTIFNAME.queue_id " (int)
112 ID of queue which this port should be mapped to.
113 .RS 7
114 .PP
115 Default:
116 .BR "None"
117 .RE
118 .SH ACTIVE-BACKUP RUNNER SPECIFIC OPTIONS
119 .TP
120 .BR "runner.hwaddr_policy " (string)
121 This defines the policy of how hardware addresses of team device and port devices should be set during the team lifetime. The following are available:
122 .RS 7
123 .PP
124 .BR "same_all "\(em
125 All ports will always have the same hardware address as the associated team device.
126 .PP
127 .BR "by_active "\(em
128 Team device adopts the hardware address of the currently active port. This is useful when the port device is not able to change its hardware address.
129 .PP
130 .BR "only_active "\(em
131 Only the active port adopts the hardware address of the team device. The others have their own.
132 .PP
133 Default:
134 .BR "same_all"
135 .RE
136 .PP
137 .TP
138 .BR "ports.PORTIFNAME.prio " (int)
139 Port priority. The higher number means higher priority.
140 .RS 7
141 .PP
142 Default:
143 .BR "0"
144 .RE
145 .TP
146 .BR "ports.PORTIFNAME.sticky " (bool)
147 Flag which indicates if the port is sticky. If set, it means the port does not get unselected if another port with higher priority or better parameters becomes available.
148 .RS 7
149 .PP
150 Default:
151 .BR "false"
152 .RE
153 .SH LOAD BALANCE RUNNER SPECIFIC OPTIONS
154 .TP
155 .BR "runner.tx_hash " (array)
156 List of fragment types (strings) which should be used for packet Tx hash computation. The following are available:
157 .RS 7
158 .PP
159 .BR "eth "\(em
160 Uses source and destination MAC addresses.
161 .PP
162 .BR "vlan "\(em
163 Uses VLAN id.
164 .PP
165 .BR "ipv4 "\(em
166 Uses source and destination IPv4 addresses.
167 .PP
168 .BR "ipv6 "\(em
169 Uses source and destination IPv6 addresses.
170 .PP
171 .BR "ip "\(em
172 Uses source and destination IPv4 and IPv6 addresses.
173 .PP
174 .BR "l3 "\(em
175 Uses source and destination IPv4 and IPv6 addresses.
176 .PP
177 .BR "tcp "\(em
178 Uses source and destination TCP ports.
179 .PP
180 .BR "udp "\(em
181 Uses source and destination UDP ports.
182 .PP
183 .BR "sctp "\(em
184 Uses source and destination SCTP ports.
185 .PP
186 .BR "l4 "\(em
187 Uses source and destination TCP and UDP and SCTP ports.
188 .PP
189 Default:
190 .B
191 ["eth", "ipv4", "ipv6"]
192 .RE
193 .TP
194 .BR "runner.tx_balancer.name " (string)
195 Name of active Tx balancer. Active Tx balancing is disabled by default. The only value available is
196 .BR "basic".
197 .RS 7
198 .PP
199 Default:
200 .BR "None"
201 .RE
202 .TP
203 .BR "runner.tx_balancer.balancing_interval " (int)
204 In tenths of a second. Periodic interval between rebalancing.
205 .RS 7
206 .PP
207 Default:
208 .BR "50"
209 .RE
210 .SH LACP RUNNER SPECIFIC OPTIONS
211 .TP
212 .BR "runner.active " (bool)
213 If active is
214 .BR "true"
215 LACPDU frames are sent along the configured links periodically. If not, it acts as "speak when spoken to".
216 .RS 7
217 .PP
218 Default:
219 .BR "true"
220 .RE
221 .PP
222 .TP
223 .BR "runner.fast_rate " (bool)
224 Option specifies the rate at which our link partner is asked to transmit LACPDU packets. If this is
225 .BR "true"
226 then packets will be sent once per second. Otherwise they will be sent every 30 seconds.
227 .RS 7
228 .PP
229 Default:
230 .BR "false"
231 .RE
232 .TP
233 .BR "runner.tx_hash " (array)
234 Same as for load balance runner.
235 .TP
236 .BR "runner.tx_balancer.name " (string)
237 Same as for load balance runner.
238 .TP
239 .BR "runner.tx_balancer.balancing_interval " (int)
240 Same as for load balance runner.
241 .TP
242 .BR "runner.sys_prio " (int)
243 System priority, value can be 0 \(en 65535.
244 .RS 7
245 .PP
246 Default:
247 .BR "65535"
248 .RE
249 .TP
250 .BR "runner.min_ports " (int)
251 Specifies the minimum number of ports that must be active before asserting carrier in the master interface, value can be 1 \(en 255.
252 .RS 7
253 .PP
254 Default:
255 .BR "1"
256 .RE
257 .TP
258 .BR "runner.agg_select_policy " (string)
259 This selects the policy of how the aggregators will be selected. The following are available:
260 .RS 7
261 .PP
262 .BR "lacp_prio "\(em
263 Aggregator with highest priority according to LACP standard will be selected. Aggregator priority is affected by per-port option
264 .BR "lacp_prio".
265 .PP
266 .BR "lacp_prio_stable "\(em
267 Same as previous one, except do not replace selected aggregator if it is still usable.
268 .PP
269 .BR "bandwidth "\(em
270 Select aggregator with highest total bandwidth.
271 .PP
272 .BR "count "\(em
273 Select aggregator with highest number of ports.
274 .PP
275 .BR "port_config "\(em
276 Aggregator with highest priority according to per-port options
277 .BR "prio " and
278 .BR "sticky "
279 will be selected. This means that the aggregator containing the port with the highest priority will be selected unless at least one of the ports in the currently selected aggregator is sticky.
280 .PP
281 Default:
282 .BR "lacp_prio"
283 .RE
284 .TP
285 .BR "ports.PORTIFNAME.lacp_prio " (int)
286 Port priority according to LACP standard. The lower number means higher priority.
287 .RS 7
288 .PP
289 Default:
290 .BR "255"
291 .RE
292 .TP
293 .BR "ports.PORTIFNAME.lacp_key " (int)
294 Port key according to LACP standard. It is only possible to aggregate ports with the same key.
295 .RS 7
296 .PP
297 Default:
298 .BR "0"
299 .RE
300 .PP
301 .SH ETHTOOL LINK WATCH SPECIFIC OPTIONS
302 .TP
303 .BR "link_watch.delay_up "| " ports.PORTIFNAME.link_watch.delay_up " (int)
304 Value is a positive number in milliseconds. It is the delay between the link coming up and the runner being notified about it.
305 .RS 7
306 .PP
307 Default:
308 .BR "0"
309 .RE
310 .TP
311 .BR "link_watch.delay_down "| " ports.PORTIFNAME.link_watch.delay_down " (int)
312 Value is a positive number in milliseconds. It is the delay between the link going down and the runner being notified about it.
313 .RS 7
314 .PP
315 Default:
316 .BR "0"
317 .RE
318 .PP
319 .SH ARP PING LINK WATCH SPECIFIC OPTIONS
320 .TP
321 .BR "link_watch.interval "| " ports.PORTIFNAME.link_watch.interval " (int)
322 Value is a positive number in milliseconds. It is the interval between ARP requests being sent.
323 .RS 7
324 .PP
325 Default:
326 .BR "1000"
327 .RE
328 .TP
329 .BR "link_watch.init_wait "| " ports.PORTIFNAME.link_watch.init_wait " (int)
330 Value is a positive number in milliseconds. It is the delay between link watch initialization and the first ARP request being sent.
331 .RS 7
332 .PP
333 Default:
334 .BR "0"
335 .RE
336 .TP
337 .BR "link_watch.missed_max "| " ports.PORTIFNAME.link_watch.missed_max " (int)
338 Maximum number of missed ARP replies. If this number is exceeded, link is reported as down.
339 .RS 7
340 .PP
341 Default:
342 .BR "3"
343 .RE
344 .TP
345 .BR "link_watch.source_host "| " ports.PORTIFNAME.link_watch.source_host " (hostname)
346 Hostname to be converted to IP address which will be filled into ARP request as source address.
347 .RS 7
348 .PP
349 Default:
350 .BR "0.0.0.0"
351 .RE
352 .TP
353 .BR "link_watch.target_host "| " ports.PORTIFNAME.link_watch.target_host " (hostname)
354 Hostname to be converted to IP address which will be filled into ARP request as destination address.
355 .TP
356 .BR "link_watch.validate_active "| " ports.PORTIFNAME.link_watch.validate_active " (bool)
357 Validate received ARP packets on active ports. If this is not set, all incoming ARP packets will be considered as a good reply.
358 .RS 7
359 .PP
360 Default:
361 .BR "false"
362 .RE
363 .TP
364 .BR "link_watch.validate_inactive "| " ports.PORTIFNAME.link_watch.validate_inactive " (bool)
365 Validate received ARP packets on inactive ports. If this is not set, all incoming ARP packets will be considered as a good reply.
366 .RS 7
367 .PP
368 Default:
369 .BR "false"
370 .RE
371 .TP
372 .BR "link_watch.vlanid "| " ports.PORTIFNAME.link_watch.vlanid " (int)
373 By default, ARP requests are sent without VLAN tags. This option causes outgoing ARP requests to be sent with the specified VLAN ID number.
374 .RS 7
375 .PP
376 Default:
377 .BR "None"
378 .RE
379 .TP
380 .BR "link_watch.send_always "| " ports.PORTIFNAME.link_watch.send_always " (bool)
381 By default, ARP requests are sent on active ports only. This option allows sending even on inactive ports.
382 .RS 7
383 .PP
384 Default:
385 .BR "false"
386 .PP
387 .SH NS/NA PING LINK WATCH SPECIFIC OPTIONS
388 .TP
389 .BR "link_watch.interval "| " ports.PORTIFNAME.link_watch.interval " (int)
390 Value is a positive number in milliseconds. It is the interval between sending NS packets.
391 .RS 7
392 .PP
393 Default:
394 .BR "1000"
395 .RE
396 .TP
397 .BR "link_watch.init_wait "| " ports.PORTIFNAME.link_watch.init_wait " (int)
398 Value is a positive number in milliseconds. It is the delay between link watch initialization and the first NS packet being sent.
399 .TP
400 .BR "link_watch.missed_max "| " ports.PORTIFNAME.link_watch.missed_max " (int)
401 Maximum number of missed NA reply packets. If this number is exceeded, link is reported as down.
402 .RS 7
403 .PP
404 Default:
405 .BR "3"
406 .RE
407 .TP
408 .BR "link_watch.target_host "| " ports.PORTIFNAME.link_watch.target_host " (hostname)
409 Hostname to be converted to IPv6 address which will be filled into NS packet as target address.
410 .SH EXAMPLES
411 .PP
412 .nf
413 {
414 "device": "team0",
415 "runner": {"name": "roundrobin"},
416 "ports": {"eth1": {}, "eth2": {}}
417 }
418 .fi
419 .PP
420 Very basic configuration.
421 .PP
422 .nf
423 {
424 "device": "team0",
425 "runner": {"name": "activebackup"},
426 "link_watch": {"name": "ethtool"},
427 "ports": {
428 "eth1": {
429 "prio": \-10,
430 "sticky": true
431 },
432 "eth2": {
433 "prio": 100
434 }
435 }
436 }
437 .fi
438 .PP
439 This configuration uses active-backup runner with ethtool link watcher. Port eth2 has higher priority, but the sticky flag ensures that if eth1 becomes active, it stays active while the link remains up.
440 .PP
441 .nf
442 {
443 "device": "team0",
444 "runner": {"name": "activebackup"},
445 "link_watch": {
446 "name": "ethtool",
447 "delay_up": 2500,
448 "delay_down": 1000
449 },
450 "ports": {
451 "eth1": {
452 "prio": \-10,
453 "sticky": true
454 },
455 "eth2": {
456 "prio": 100
457 }
458 }
459 }
460 .fi
461 .PP
462 Similar to the previous one. Only difference is that link changes are not propagated to the runner immediately, but delays are applied.
463 .PP
464 .nf
465 {
466 "device": "team0",
467 "runner": {"name": "activebackup"},
468 "link_watch": {
469 "name": "arp_ping",
470 "interval": 100,
471 "missed_max": 30,
472 "target_host": "192.168.23.1"
473 },
474 "ports": {
475 "eth1": {
476 "prio": \-10,
477 "sticky": true
478 },
479 "eth2": {
480 "prio": 100
481 }
482 }
483 }
484 .fi
485 .PP
486 This configuration uses ARP ping link watch.
487 .PP
488 .nf
489 {
490 "device": "team0",
491 "runner": {"name": "activebackup"},
492 "link_watch": [
493 {
494 "name": "arp_ping",
495 "interval": 100,
496 "missed_max": 30,
497 "target_host": "192.168.23.1"
498 },
499 {
500 "name": "arp_ping",
501 "interval": 50,
502 "missed_max": 20,
503 "target_host": "192.168.24.1"
504 }
505 ],
506 "ports": {
507 "eth1": {
508 "prio": \-10,
509 "sticky": true
510 },
511 "eth2": {
512 "prio": 100
513 }
514 }
515 }
516 .fi
517 .PP
518 Similar to the previous one, only this time two link watchers are used at the same time.
519 .PP
520 .nf
521 {
522 "device": "team0",
523 "runner": {
524 "name": "loadbalance",
525 "tx_hash": ["eth", "ipv4", "ipv6"]
526 },
527 "ports": {"eth1": {}, "eth2": {}}
528 }
529 .fi
530 .PP
531 Configuration for hash-based passive Tx load balancing.
532 .PP
533 .nf
534 {
535 "device": "team0",
536 "runner": {
537 "name": "loadbalance",
538 "tx_hash": ["eth", "ipv4", "ipv6"],
539 "tx_balancer": {
540 "name": "basic"
541 }
542 },
543 "ports": {"eth1": {}, "eth2": {}}
544 }
545 .fi
546 .PP
547 Configuration for active Tx load balancing using basic load balancer.
548 .PP
549 .nf
550 {
551 "device": "team0",
552 "runner": {
553 "name": "lacp",
554 "active": true,
555 "fast_rate": true,
556 "tx_hash": ["eth", "ipv4", "ipv6"]
557 },
558 "link_watch": {"name": "ethtool"},
559 "ports": {"eth1": {}, "eth2": {}}
560 }
561 .fi
562 .PP
563 Configuration for connection to LACP capable counterpart.
564 .SH SEE ALSO
565 .BR teamd (8),
566 .BR teamdctl (8),
567 .BR teamnl (8),
568 .BR bond2team (1)
569 .SH AUTHOR
570 .PP
571 Jiri Pirko is the original author and current maintainer of libteam.