Description of problem:
When adding a new knet link, these two error messages appear in corosync.log:
1) knet_link_set_ping_timers for nodeid 1, link 1 failed: Invalid argument (22)
2) New config has different knet transport for link 1. Internal value was NOT changed. To reconfigure an interface it must be deleted and recreated. A working interface needs to be available to corosync at all times

The cluster is running just fine, though.


Version-Release number of selected component (if applicable):
corosync-3.0.0-2.el8.x86_64
corosynclib-3.0.0-2.el8.x86_64
corosync-qdevice-3.0.0-2.el8.x86_64
libknet1-1.4-3.el8.x86_64
libknet1-compress-bzip2-plugin-1.4-3.el8.x86_64
libknet1-compress-lz4-plugin-1.4-3.el8.x86_64
libknet1-compress-lzma-plugin-1.4-3.el8.x86_64
libknet1-compress-lzo2-plugin-1.4-3.el8.x86_64
libknet1-compress-plugins-all-1.4-3.el8.x86_64
libknet1-compress-zlib-plugin-1.4-3.el8.x86_64
libknet1-crypto-nss-plugin-1.4-3.el8.x86_64
libknet1-crypto-openssl-plugin-1.4-3.el8.x86_64
libknet1-crypto-plugins-all-1.4-3.el8.x86_64
libknet1-plugins-all-1.4-3.el8.x86_64


How reproducible:
always, easily


Steps to Reproduce:
1. Add a new knet link.
2. Check corosync.log.
3. Delete a knet link.
4. Add a link with the same linknumber as the link deleted in step 3 had.
2. Check corosync.log.


Actual results:
When adding a link for the first time, both messages 1) and 2) from above get logged. When adding a previously deleted link, only message 1) is logged.


Expected results:
The messages seem to be false since the cluster is running fine. It is therefore expected for them not to be logged.



Additional info:

corosync.conf before adding a link:
totem {
     version: 2
     cluster_name: rhel80-knet
     transport: knet
     crypto_cipher: aes256
     crypto_hash: sha256
}

nodelist {
     node {
         name: rh80-node1
         nodeid: 1
         ring0_addr: 192.168.122.201
     }

     node {
         name: rh80-node2
         nodeid: 2
         ring0_addr: 192.168.122.202
     }
}

quorum {
     provider: corosync_votequorum
     two_node: 1
}

logging {
     to_logfile: yes
     logfile: /var/log/cluster/corosync.log
     to_syslog: yes
}



corosync.conf after adding a link:
totem {
     version: 2
     cluster_name: rhel80-knet
     transport: knet
     crypto_cipher: aes256
     crypto_hash: sha256
}

nodelist {
     node {
         name: rh80-node1
         nodeid: 1
         ring0_addr: 192.168.122.201
         ring1_addr: 192.168.123.201
     }

     node {
         name: rh80-node2
         nodeid: 2
         ring0_addr: 192.168.122.202
         ring1_addr: 192.168.123.202
     }
}

quorum {
     provider: corosync_votequorum
     two_node: 1
}

logging {
     to_logfile: yes
     logfile: /var/log/cluster/corosync.log
     to_syslog: yes
}



logs from node1:
Apr 30 17:42:33 [1296] rh80-node1 corosync notice  [CFG   ] Config reload requested by node 1
Apr 30 17:42:33 [1296] rh80-node1 corosync info    [TOTEM ] Configuring link 0
Apr 30 17:42:33 [1296] rh80-node1 corosync info    [TOTEM ] Configured link number 0: local addr: 192.168.122.201, port=5405
Apr 30 17:42:33 [1296] rh80-node1 corosync info    [TOTEM ] Configuring link 1
Apr 30 17:42:33 [1296] rh80-node1 corosync info    [TOTEM ] Configured link number 1: local addr: 192.168.123.201, port=5406
Apr 30 17:42:33 [1296] rh80-node1 corosync error   [TOTEM ] New config has different knet transport for link 1. Internal value was NOT changed.
Apr 30 17:42:33 [1296] rh80-node1 corosync error   [TOTEM ] To reconfigure an interface it must be deleted and recreated. A working interface needs to be available to corosync at all times
Apr 30 17:42:33 [1296] rh80-node1 corosync error   [TOTEM ] knet_link_set_ping_timers for nodeid 2, link 1 failed: Invalid argument (22)
Apr 30 17:42:34 [1296] rh80-node1 corosync info    [KNET  ] rx: host: 2 link: 1 is up
Apr 30 17:42:34 [1296] rh80-node1 corosync info    [KNET  ] pmtud: PMTUD link change for host: 2 link: 1 from 470 to 1366



logs from node2:
Apr 30 17:42:33 [10063] rh80-node2 corosync notice  [CFG   ] Config reload requested by node 1
Apr 30 17:42:33 [10063] rh80-node2 corosync info    [TOTEM ] Configuring link 0
Apr 30 17:42:33 [10063] rh80-node2 corosync info    [TOTEM ] Configured link number 0: local addr: 192.168.122.202, port=5405
Apr 30 17:42:33 [10063] rh80-node2 corosync info    [TOTEM ] Configuring link 1
Apr 30 17:42:33 [10063] rh80-node2 corosync info    [TOTEM ] Configured link number 1: local addr: 192.168.123.202, port=5406
Apr 30 17:42:33 [10063] rh80-node2 corosync error   [TOTEM ] New config has different knet transport for link 1. Internal value was NOT changed.
Apr 30 17:42:33 [10063] rh80-node2 corosync error   [TOTEM ] To reconfigure an interface it must be deleted and recreated. A working interface needs to be available to corosync at all times
Apr 30 17:42:33 [10063] rh80-node2 corosync error   [TOTEM ] knet_link_set_ping_timers for nodeid 1, link 1 failed: Invalid argument (22)
Apr 30 17:42:34 [10063] rh80-node2 corosync info    [KNET  ] rx: host: 1 link: 1 is up
Apr 30 17:42:34 [10063] rh80-node2 corosync info    [KNET  ] pmtud: PMTUD link change for host: 1 link: 1 from 470 to 1366