In your host OS and multipath software, set the time for monitoring recovery from a failure. If an alternate LU path fails during reboot for firmware update or during reset due to a failure, both the original path and alternate path become temporarily inaccessible until the reboot or reset is completed. A 25-Gbps iSCSI channel board requires approximately 130 seconds at maximum before I/O restart for reboot during update of its firmware or for reset if a failure occurs. To continue host I/O during this time, use the multipath configuration to connect storage systems to hosts.
I/O might stop in the following circumstances:
- If the single path configuration is used, host I/O might stop, or a host I/O error might occur because the 25-Gbps iSCSI channel board cannot continue host I/O for a maximum of 130 seconds during reboot for firmware update or during reset due to a failure.
- If the parameter for monitoring recovery from a failure is not configured and an alternate LU path fails during reboot for firmware update or during reset due to a failure, I/O might stop or a host I/O error might occur.
The relevant parameter must be configured for the following host OSes and multipath software:
- Windows Server 2019 and 2022 (Microsoft DSM/MPIO
- Red Hat Enterprise Linux (Device Mapper Multipath
- SUSE Linux (Device Mapper Multipath
If VMware uses a guest OS iSCSI initiator, you need to configure the relevant parameter in the guest OS (Windows, Red Hat Enterprise Linux, or SUSE Linux).
When you use VMware ESXi as the iSCSI initiator, you do not need to configure the parameter.