Skip to contentSkip to conversation

bonding: fix active-backup failover for current ARP slave

PATCH In review Jiri Wiesner 2 participants 2 messages Aug 16, 2020 – Aug 18, 2020

Subsystems (3)
lore
Share & tools
lore
#0 lore raw

PATCH 1/1 bonding: fix active-backup failover for current ARP slave

When the ARP monitor is used for link detection, ARP replies are validated for all slaves (arp_validate=3) and fail_over_mac is set to active, two slaves of an active-backup bond may get stuck in a state where both of them are active and pass packets that they receive to the bond. This state makes IPv6 duplicate address detection fail. The state is reached thus:
1. The current active slave goes down because the ARP target is not reachable.
2. The current ARP slave is chosen and made active.
3. A new slave is enslaved. This new slave becomes the current active slave and can reach the ARP target.
As a result, the current ARP slave stays active after the enslave action has finished and the log is littered with "PROBE BAD" messages:

Quoted 1 line

bond0: PROBE: c_arp ens10 && cas ens11 BAD

The workaround is to remove the slave with "going back" status from the bond and re-enslave it. This issue was encountered when DPDK PMD interfaces were being enslaved to an active-backup bond.

I would be possible to fix the issue in bond_enslave() or bond_change_active_slave() but the ARP monitor was fixed instead to keep most of the actions changing the current ARP slave in the ARP monitor code. The current ARP slave is set as inactive and backup during the commit phase. A new state, BOND_LINK_FAIL, has been introduced for slaves in the context of the ARP monitor. This allows administrators to see how slaves are rotated for sending ARP requests and attempts are made to find a new active slave.

Fixes b2220cad583c9 ("bonding: refactor ARP active-backup monitor")
---
 drivers/net/bonding/bond_main.c | 18 ++++++++++++++++--
 1 file changed, 16 insertions(+), 2 deletions(-)
drivers/net/bonding/bond_main.c +16 -2
index c853ca67058c..0ee59ea357f5 100644
@@ -2945,6 +2945,9 @@ static int bond_ab_arp_inspect(struct bonding *bond)
 			if (bond_time_in_interval(bond, last_rx, 1)) {
 				bond_propose_link_state(slave, BOND_LINK_UP);
 				commit++;
+			} else if (slave->link == BOND_LINK_BACK) {
+				bond_propose_link_state(slave, BOND_LINK_FAIL);
+				commit++;
 			}
 			continue;
 		}
@@ -3053,6 +3056,19 @@ static void bond_ab_arp_commit(struct bonding *bond)
 
 			continue;
 
+		case BOND_LINK_FAIL:
+			bond_set_slave_link_state(slave, BOND_LINK_FAIL,
+						  BOND_SLAVE_NOTIFY_NOW);
+			bond_set_slave_inactive_flags(slave,
+						      BOND_SLAVE_NOTIFY_NOW);
+
+			/* A slave has just been enslaved and has become
+			 * the current active slave.
+			 */
+			if (rtnl_dereference(bond->curr_active_slave))
+				RCU_INIT_POINTER(bond->current_arp_slave, NULL);
+			continue;
+
 		default:
 			slave_err(bond->dev, slave->dev,
 				  "impossible: link_new_state %d on slave\n",
@@ -3103,8 +3119,6 @@ static bool bond_ab_arp_probe(struct bonding *bond)
 			return should_notify_rtnl;
 	}
 
-	bond_set_slave_inactive_flags(curr_arp_slave, BOND_SLAVE_NOTIFY_LATER);
-
 	bond_for_each_slave_rcu(bond, slave, iter) {
 		if (!found && !before && bond_slave_is_up(slave))
 			before = slave;
Signature · 1 line
2.26.2
maintainer
#1 lore raw

From: Jiri Wiesner <[email protected]>
Date: Sun, 16 Aug 2020 20:52:44 +0200

Quoted 27 lines

When the ARP monitor is used for link detection, ARP replies are validated for all slaves (arp_validate=3) and fail_over_mac is set to active, two slaves of an active-backup bond may get stuck in a state where both of them are active and pass packets that they receive to the bond. This state makes IPv6 duplicate address detection fail. The state is reached thus:
1. The current active slave goes down because the ARP target is not reachable.
2. The current ARP slave is chosen and made active.
3. A new slave is enslaved. This new slave becomes the current active slave and can reach the ARP target.
As a result, the current ARP slave stays active after the enslave action has finished and the log is littered with "PROBE BAD" messages:

Quoted 1 line

bond0: PROBE: c_arp ens10 && cas ens11 BAD

The workaround is to remove the slave with "going back" status from the bond and re-enslave it. This issue was encountered when DPDK PMD interfaces were being enslaved to an active-backup bond.

I would be possible to fix the issue in bond_enslave() or bond_change_active_slave() but the ARP monitor was fixed instead to keep most of the actions changing the current ARP slave in the ARP monitor code. The current ARP slave is set as inactive and backup during the commit phase. A new state, BOND_LINK_FAIL, has been introduced for slaves in the context of the ARP monitor. This allows administrators to see how slaves are rotated for sending ARP requests and attempts are made to find a new active slave.

Quoted 2 lines
Fixes b2220cad583c9 ("bonding: refactor ARP active-backup monitor")

Applied and queued up for -stable, thanks Jiri.