TEAMD.CONF(5)

📦
version 1.32 · 2 versions
, in arch rolling, fedora 43, fedora 44, opensuse-leap 16.0

TEAMD.CONF(5) — team daemon configuration file 💖

okay bestie, welcome to the configuration file for libteam's daemon — this is teamd.conf, the whole vibe manual, section 5 of the man pages. it's a lot, i know. but you're gonna eat this up, fr fr. 🎀

✨ NAME ✨

teamd.conf — the libteam daemon configuration file. yeah. that's the whole name. iconic.

🌸 DESCRIPTION 🌸

teamd uses JSON format configuration. that's it, that's the sentence. JSON. curly braces. commas you WILL forget. slay. 💅

🦋 OPTIONS 🦋

  • device (string) — the desired name of your new team device. name it something cute. or don't. it's your team. ✨
  • debug_level (int) — how loud the debug messages get. higher number = more debug messages printed. literally the same thing as tacking on -g command line options.

Default: 0 (disabled)

  • hwaddr (string) — the desired hardware address of your new team device. the usual MAC address format is accepted, so go off. 🌈
  • runner.name (string) — the name of the team device's runner. here's the whole roster, bestie:
    • broadcast — a simple runner that directs the team device to transmit packets via all ports. everyone gets a packet. no port left behind. 🫶
    • roundrobin — a simple runner that directs the team device to transmit packets in a round-robin fashion. taking turns, very polite.
    • random — a simple runner that directs the team device to transmit packets on a randomly selected port. chaos mode. 🎲
    • activebackup — watches for link changes and selects the active port to be used for data transfers. main character port energy. 👑
    • loadbalance — for passive load balancing, the runner only sets up a BPF hash function which will determine the port for packet transmit. for active load balancing, the runner moves hashes among available ports trying to reach perfect balance. it's giving balance. ⚖️
    • lacp — implements the 802.3ad LACP protocol. can use the same Tx port selection possibilities as the loadbalance runner.

Default: roundrobin

  • notify_peers.count (int) — how many bursts of unsolicited NAs and gratuitous ARP packets get sent after a port is enabled or disabled.

Default: 0 (disabled)

Default for activebackup runner: 1

  • notify_peers.interval (int) — a positive number in milliseconds. specifies the interval between bursts of notify-peer packets.

Default: 0

  • mcast_rejoin.count (int) — how many bursts of multicast group rejoin requests get sent after a port is enabled or disabled.

Default: 0 (disabled)

Default for activebackup runner: 1

  • mcast_rejoin.interval (int) — a positive number in milliseconds. specifies the interval between bursts of multicast group rejoin requests.

Default: 0

  • link_watch.name | ports.PORTIFNAME.link_watch.name (string) — the name of the link watcher to be used. the available link watchers are:
    • ethtool — uses the Libteam lib to get port ethtool state changes. lowkey the chill one.
    • arp_ping — ARP requests are sent through a port. if an ARP reply is received, the link is considered to be up. it's literally asking "u up?" 💌
    • nsna_ping — similar to the previous one, except it uses the IPv6 Neighbor Solicitation / Neighbor Advertisement mechanism. this is an alternative to arp_ping and becomes handy in pure-IPv6 environments. ipv6 girlies rise 🌐
  • ports (object) — the list of ports, i.e. network devices, to be used in a team device.

See examples for more information. (yes bestie, examples exist. we'll get there. 🫶)

  • ports.PORTIFNAME.queue_id (int) — the ID of the queue which this port should be mapped to.

Default: None

💜 ACTIVE-BACKUP RUNNER SPECIFIC OPTIONS 💜

  • runner.hwaddr_policy (string) — this defines the policy of how hardware addresses of the team device and port devices should be set during the team lifetime. here's what's available:
    • same_all — all ports will always have the same hardware address as the associated team device. matching outfits. 🎀
    • by_active — the 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.
    • only_active — only the active port adopts the hardware address of the team device. the others have their own. individuality 💅

Default: same_all

  • ports.PORTIFNAME.prio (int) — port priority. the higher number means higher priority.

Default: 0

  • ports.PORTIFNAME.sticky (bool) — a flag that says whether the port is sticky. if it's set, it means the port does not get unselected if another port with higher priority or better parameters becomes available. it's loyal. it stays. 😤

Default: false

🔥 LOAD BALANCE RUNNER SPECIFIC OPTIONS 🔥

  • runner.tx_hash (array) — a list of fragment types (strings) which should be used for packet Tx hash computation. here's the lineup:
    • eth — uses source and destination MAC addresses.
    • vlan — uses VLAN id.
    • ipv4 — uses source and destination IPv4 addresses.
    • ipv6 — uses source and destination IPv6 addresses.
    • ip — uses source and destination IPv4 and IPv6 addresses.
    • l3 — uses source and destination IPv4 and IPv6 addresses.
    • tcp — uses source and destination TCP ports.
    • udp — uses source and destination UDP ports.
    • sctp — uses source and destination SCTP ports.
    • l4 — uses source and destination TCP and UDP and SCTP ports.

Default: ["eth", "ipv4", "ipv6"]

  • runner.tx_balancer.name (string) — the name of the active Tx balancer. heads up: active Tx balancing is disabled by default. the only value available is basic.

Default: None

  • runner.tx_balancer.balancing_interval (int) — measured in tenths of a second. it's the periodic interval between rebalancing.

Default: 50

🌈 LACP RUNNER SPECIFIC OPTIONS 🌈

  • runner.active (bool) — if active is true, LACPDU frames are sent along the configured links periodically. if not, it acts as "speak when spoken to". shy but functional. 🙈

Default: true

  • runner.fast_rate (bool) — this option specifies the rate at which our link partner is asked to transmit LACPDU packets. if this is true then packets will be sent once per second. otherwise they will be sent every 30 seconds. big difference, bestie. ⏱️

Default: false

  • runner.tx_hash (array) — same as for the load balance runner.
  • runner.tx_balancer.name (string) — same as for the load balance runner.
  • runner.tx_balancer.balancing_interval (int) — same as for the load balance runner.
  • runner.sys_prio (int) — system priority. the value can be 0 – 65535.

Default: 65535

  • runner.min_ports (int) — specifies the minimum number of ports that must be active before asserting carrier in the master interface. the value can be 1 – 1024.

Default: 1

  • runner.agg_select_policy (string) — this selects the policy of how the aggregators will be selected. the options are:
    • lacp_prio — the aggregator with the highest priority according to the LACP standard will be selected. aggregator priority is affected by the per-port option lacp_prio.
    • lacp_prio_stable — same as the previous one, except it does not replace the selected aggregator if it is still usable. if it ain't broke. 🫶
    • bandwidth — select the aggregator with the highest total bandwidth.
    • count — select the aggregator with the highest number of ports. more is more.
    • port_config — the aggregator with the highest priority according to the per-port options prio and sticky 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.

Default: lacp_prio

  • ports.PORTIFNAME.lacp_prio (int) — port priority according to the LACP standard. careful: the lower number means higher priority. yes it's backwards. no i don't make the rules. 💀

Default: 255

  • ports.PORTIFNAME.lacp_key (int) — port key according to the LACP standard. it is only possible to aggregate ports with the same key.

Default: 0

  • link_watch.delay_up | ports.PORTIFNAME.link_watch.delay_up (int) — a positive number in milliseconds. it's the delay between the link coming up and the runner being notified about it.

Default: 0

  • link_watch.delay_down | ports.PORTIFNAME.link_watch.delay_down (int) — a positive number in milliseconds. it's the delay between the link going down and the runner being notified about it.

Default: 0

  • link_watch.interval | ports.PORTIFNAME.link_watch.interval (int) — a positive number in milliseconds. it's the interval between ARP requests being sent.

Default: 1000

  • link_watch.init_wait | ports.PORTIFNAME.link_watch.init_wait (int) — a positive number in milliseconds. it's the delay between link watch initialization and the first ARP request being sent.

Default: 0

  • link_watch.missed_max | ports.PORTIFNAME.link_watch.missed_max (int) — the maximum number of missed ARP replies. if this number is exceeded, the link is reported as down. 😭

Default: 3

  • link_watch.source_host | ports.PORTIFNAME.link_watch.source_host (hostname) — a hostname to be converted to an IP address, which will be filled into the ARP request as the source address.

Default: 0.0.0.0

  • link_watch.target_host | ports.PORTIFNAME.link_watch.target_host (hostname) — a hostname to be converted to an IP address, which will be filled into the ARP request as the destination address.
  • link_watch.validate_active | ports.PORTIFNAME.link_watch.validate_active (bool) — validate received ARP packets on active ports. if this is not set, all incoming ARP packets will be considered as a good reply. trusting. maybe too trusting. 🙈

Default: false

  • link_watch.validate_inactive | ports.PORTIFNAME.link_watch.validate_inactive (bool) — validate received ARP packets on inactive ports. if this is not set, all incoming ARP packets will be considered as a good reply.

Default: false

  • link_watch.vlanid | ports.PORTIFNAME.link_watch.vlanid (int) — by default, ARP requests are sent without VLAN tags. this option causes outgoing ARP requests to be sent with the specified VLAN ID number.

Default: None

  • link_watch.send_always | ports.PORTIFNAME.link_watch.send_always (bool) — by default, ARP requests are sent on active ports only. this option allows sending even on inactive ports. inclusive behavior, we love to see it. 💖

Default: false

okay bestie don't go anywhere, we are so not done — part 2 is coming and it's lowkey even more. 💅

okay bestie, this is the NS/NA ping link watcher zone — it sends Neighbor Solicitation packets and waits for Neighbor Advertisement replies. These options can live at the top level or per-port, and they're lowkey precise. 🦋

  • link_watch.interval | ports.PORTIFNAME.link_watch.interval (int) — Value is a positive number in milliseconds. It is the interval between sending NS packets. Default: 1000. ⏱️
  • link_watch.init_wait | ports.PORTIFNAME.link_watch.init_wait (int) — Value is a positive number in milliseconds. It is the delay between link watch initialization and the first NS packet being sent. 🕰️
  • link_watch.missed_max | ports.PORTIFNAME.link_watch.missed_max (int) — Maximum number of missed NA reply packets. If this number is exceeded, link is reported as down. Default: 3. 💀
  • link_watch.target_host | ports.PORTIFNAME.link_watch.target_host (hostname) — Hostname to be converted to IPv6 address which will be filled into NS packet as target address. 🎯
  • link_watch_policy (string) — Name of link-watchers evaluation policy. Available options are following:
    • any — Link is up if any of the link-watchers reports the link up.
    • all — Link is up if all of the link-watchers reports the link up.

Default: any

✨ EXAMPLES ✨

bestie, these configs are literally the blueprint. Copy them exactly, okay? 💅

{
  "device": "team0",
  "runner": {"name": "roundrobin"},
  "ports": {"eth1": {}, "eth2": {}}
}

Very basic configuration.

{
  "device": "team0",
  "runner": {"name": "activebackup"},
  "link_watch": {"name": "ethtool"},
  "ports": {
    "eth1": {
      "prio": -10,
      "sticky": true
    },
    "eth2": {
      "prio": 100
    }
  }
}

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.

{
  "device": "team0",
  "runner": {"name": "activebackup"},
  "link_watch": {
    "name": "ethtool",
    "delay_up": 2500,
    "delay_down": 1000
  },
  "ports": {
    "eth1": {
      "prio": -10,
      "sticky": true
    },
    "eth2": {
      "prio": 100
    }
  }
}

Similar to the previous one. Only difference is that link changes are not propagated to the runner immediately, but delays are applied.

{
  "device": "team0",
  "runner": {"name": "activebackup"},
  "link_watch":  {
    "name": "arp_ping",
    "interval": 100,
    "missed_max": 30,
    "target_host": "192.168.23.1"
  },
  "ports": {
    "eth1": {
      "prio": -10,
      "sticky": true
    },
    "eth2": {
      "prio": 100
    }
  }
}

This configuration uses ARP ping link watch.

{
"device": "team0",
"runner": {"name": "activebackup"},
"link_watch": [
  {
    "name": "arp_ping",
    "interval": 100,
    "missed_max": 30,
    "target_host": "192.168.23.1"
  },
  {
    "name": "arp_ping",
    "interval": 50,
    "missed_max": 20,
    "target_host": "192.168.24.1"
  }
],
"ports": {
  "eth1": {
    "prio": -10,
    "sticky": true
  },
  "eth2": {
    "prio": 100
    }
  }
}

Similar to the previous one, only this time two link watchers are used at the same time.

{
  "device": "team0",
  "runner": {"name": "activebackup"},
  "link_watch_policy": "all",
  "link_watch": [
    {
    "name": "arp_ping",
    "interval": 100,
    "missed_max": 30,
    "target_host": "192.168.23.1"
    },
    {
    "name": "ethtool"
    }
  ],
  "ports":  {
    "eth1": {
      "prio": -10,
      "sticky": true
    },
    "eth2": {
      "prio": 100
    }
  }
}

Two link-watchers are used at the same time. Link is up only if all configured link-watchers report link is up.

{
  "device": "team0",
  "runner": {
    "name": "loadbalance",
    "tx_hash": ["eth", "ipv4", "ipv6"]
  },
  "ports": {"eth1": {}, "eth2": {}}
}

Configuration for hash-based passive Tx load balancing.

{
  "device": "team0",
  "runner": {
    "name": "loadbalance",
    "tx_hash": ["eth", "ipv4", "ipv6"],
    "tx_balancer": {
      "name": "basic"
    }
  },
  "ports": {"eth1": {}, "eth2": {}}
}

Configuration for active Tx load balancing using basic load balancer.

{
  "device": "team0",
  "runner": {
    "name": "lacp",
    "active": true,
    "fast_rate": true,
    "tx_hash": ["eth", "ipv4", "ipv6"]
  },
  "link_watch": {"name": "ethtool"},
  "ports": {"eth1": {}, "eth2": {}}
}

Configuration for connection to LACP capable counterpart.

✨ SEE ALSO ✨

✨ AUTHOR ✨

Jiri Pirko is the original author and current maintainer of libteam. 👑

This page is from libteam, dated 2013-07-09, and it's the TEAMD.CONF(5) man page. That's the source/version and date, bestie — slay. 💖

📦 where this page comes from