30 seconds of ping loss with a single-mode ifgrp during a switch OS upgrade
Applies to
- ONTAP 9
- Single mode ifgrp
Issue
- This is a single-mode ifgrp configuration
ifgrp: a0amember ports: e1a, e1bCrate policy: singlemode- During the switch OS upgrade, an ifgrp member port e1a flapped (down/up), resulting in approximately 30 seconds of PING loss between the client and the storage system
- If the e1a port goes down, the new port status will be detected by the Link Monitoring logic, which checks port status at 5-second intervals. Then, subsequent traffic will be switched to the e1b port
[node-01: kernel: netif.linkDown:info]: Ethernet e1a: Link down, check cable.[node-01: vifmgr: vifmgr.portdown:notice]: A link down event was received on node node-01, port e1a.[node-01: vifmgr: vifmgr.reach.noreach:notice]: Network port e1a on node node-01 cannot reach its expected broadcast domain Default:service. No other broadcast domains appear to be reachable from this port.[node-01: vifmgr: vifmgr.reach.skipped:notice]: Network port e1a on node node-01 was not scanned for reachability because it was administratively or operationally down at the time of the scan.- After the e1a port comes back up, the new port status will be detected by the Link Monitoring logic again, and then communication will switch back to that port
[node-01: kernel: netif.linkUp:info]: Ethernet e1a: Link up.[node-01: vifmgr: vifmgr.portup:notice]: A link up event was received on node node-01, port e1a.[node-01: vifmgr: vifmgr.reach.ok:notice]: Network port e1a on node node-01 can reach its expected broadcast domain Default:service. No other broadcast domains appear to be reachable from this port.