Corosync Crash

[Sig]

Hallo !
Wir haben einen Testcluster mit Proxmox VE 6.0-6 gebaut, und Corosync scheint nicht sonderlich stabil zu sein.
Gerade eben ist ein Node aus dem cluster herausgefallen. (Corosync ist nicht mehr gelaufen)

 

[Sig]

corosync lies sich aber mit "service restart ..." wieder starten.

service corosync status
● corosync.service - Corosync Cluster Engine
Loaded: loaded (/lib/systemd/system/corosync.service; enabled; vendor preset: enabled)
Active: failed (Result: signal) since Thu 2019-09-05 15:00:13 CEST; 46min ago
Docs: man:corosync
man:corosync.conf
man:corosync_overview
Process: 1285 ExecStart=/usr/sbin/corosync -f $COROSYNC_OPTIONS (code=killed, signal=FPE)
Main PID: 1285 (code=killed, signal=FPE)
Sep 05 15:00:10 proxmox-hyp-01 corosync[1285]: [KNET ] host: host: 2 has no active links
Sep 05 15:00:10 proxmox-hyp-01 corosync[1285]: [KNET ] link: host: 1 link: 0 is down
Sep 05 15:00:10 proxmox-hyp-01 corosync[1285]: [KNET ] host: host: 1 (passive) best link: 0 (pri: 1)
Sep 05 15:00:10 proxmox-hyp-01 corosync[1285]: [KNET ] host: host: 1 has no active links
Sep 05 15:00:12 proxmox-hyp-01 corosync[1285]: [KNET ] rx: host: 2 link: 0 is up
Sep 05 15:00:12 proxmox-hyp-01 corosync[1285]: [KNET ] rx: host: 1 link: 0 is up
Sep 05 15:00:12 proxmox-hyp-01 corosync[1285]: [KNET ] host: host: 2 (passive) best link: 0 (pri: 1)
Sep 05 15:00:12 proxmox-hyp-01 corosync[1285]: [KNET ] host: host: 1 (passive) best link: 0 (pri: 1)
Sep 05 15:00:13 proxmox-hyp-01 systemd[1]: corosync.service: Main process exited, code=killed, status=8/FPE
Sep 05 15:00:13 proxmox-hyp-01 systemd[1]: corosync.service: Failed with result 'signal'.

 

[Sig]

dmesg :
[Thu Sep 5 15:00:16 2019] traps: corosync[1299] trap divide error ip:7fb7dbf54f56 sp:7fb7d071fa50 error:0 in libknet.so.1.3.0[7fb7dbf49000+13000]



ii corosync 3.0.2-pve2 amd64 cluster engine daemon and utilities
ii libcorosync-common4:amd64 3.0.2-pve2 amd64 cluster engine common library
ii libknet1:amd64 1.11-pve1 amd64 kronosnet core switching implementation


Zusatzinfos :
Die Nodes sind über ein 10G (active-backup Bonding) Netzwerk miteinenader verbunden.