[LU-9480] LNet Dynamic Discovery

• Type: New Feature
• Resolution: Fixed
• Priority: Minor
• Fix Version/s: Lustre 2.11.0
• Assignee: Amir Shehata (Inactive)
• Reporter: Amir Shehata (Inactive)

Description

This ticket will track Dynamic Discovery feature

lnet: show peer state
lnet: Fix deleting peers from YAML
lnet: add enhanced statistics
lnet: selftest MR fix
lnet: add "lnetctl discover"
lnet: fix for static analysis issues
lnet: add "lnetctl ping" command
lnet: add "lnetctl peer list"
lnet: implement Peer Discovery
lnet: add the Push target
lnet: tune lnet_peer_discovery_disabled with lnetctl
lnet: add discovery thread
lnet: add msg_type to lnet_event
lnet: reference counts on lnet_peer/lnet_peer_net
lnet: add LNET_PEER_CONFIGURED flag
lnet: preferred NIs for non-Multi-Rail peers
lnet: introduce LNET_PEER_MULTI_RAIL flag bit
lnet: refactor lnet_add_peer_ni()
lnet: refactor lnet_del_peer_ni()
lnet: rename lnet_add/del_peer_ni_to/from_peer()
lnet: cleanup of lnet_peer_ni_addref/decref_locked()
lnet: add sanity checks on ping-related constants
lnet: add Multi-Rail and Discovery ping feature bits
lnet: automatic sizing of router pinger buffers
lnet: add struct lnet_ping_buffer
lnet: remove debug ioctl
lnet: configure lnet_max_interfaces tunable from dlc
lnet: add lnet_max_interfaces tunable

Issue Links

• is duplicated by: LU-6531 Fujitsu's o2iblnd Channel Bonding Solution
• is related to: LU-9914 Dynamic Discovery - discovery hangs if max_interfaces is changed from 200->16
• is related to: LU-17700 lnet ping: failed to ping NID: Protocol error
• is related to: LU-9077 New static analysis issues in v2_9_52_0-66-gec839d4
• is related to: LU-9904 LNetError: 16885:0:(peer.c:1786:lnet_peer_push_event()) Push Put from unknown 0@<0:0> (source 0@<0:0>)
• is related to: LU-9911 lnet_peer_discovery_complete() must use safe list iteration
• is related to: LU-7734 LNet Multi-Rail Project
• is related to: LU-7562 Increase LNET_MAX_INTERFACES