Hosts loses redundant IPs from iSCSI storage

[SteBre]

Affected version

• oVirt Engine: 4.5.5-1.22.el8
• Host OS Variant: RHEL - 8.10 - 1.0.7.el8
• VDSM version: vdsm-4.50.5.1-4.el8

Describe the bug

We have 2 Hosts in one Cluster, both attached with 8 IPs to a HA-iSCSI Storage. After rebooting a host, the attached iSCSI storage will connect only 2 of 8 IPs. As a result, there are of course many error messages from iSCSI storage that the connections are no longer redundant.

the missing 6 IP addresses then have to be reconnected. then it works again until the next reboot.

To reproduce

Reboot one of the hosts.

Expected behavior

All connected IPs should be connected again.

Screenshots

After rebooting only IPs .21 and .22 are connected.

[dupondje]

If you go to the connections of the datastore, how many connections are shown there?

[SteBre]

there are only 2 connections shown: