White Papers

Server configuration
9 Dell EMC SC Series: Red Hat Enterprise Linux Best Practices | CML1031
Queue depth settings for a specific Emulex HBA:
# cat /sys/class/scsi_host/host12/lpfc_lun_queue_depth
# cat /sys/class/scsi_host/host12/lpfc_hba_queue_depth
Note: the HBA settings might not match the global settings if the HBA does not support them.
To change the HBA queue depth settings, include the following line to the HBA driver module configuration
file (section 2.1) and reload the driver module (section 2.1.3).
options qla2xxx ql2xmaxqdepth=<value> #For Qlogic
options lpfc lpfc_lun_queue_depth=<value> lpfc_hba_queue_depth=<value> #For
Emulex
2.1.2 Setting the HBA port timeout
During an SC Series path or controller failover while operating in legacy port mode, failing ports trigger
their respective World Wide Port Name (WWPN) identities to momentarily disappear from the SAN fabric. The
WWPN identities then relocate to a reserve port within the same fault domain on the alternate active
controller.
During a path failover while operating in virtual port mode, the WWPN identities of failing N_Port ID
virtualization (NPIV) ports are relocated to another active NPIV port within the same fault domain on the same
controller. During a controller failover, the WWPN identities of any failing NPIV ports are relocated to the
active NPIV ports within the same fault domain on the alternate active controller.
In either failover scenario, the SC Series storage system may take up to 60 seconds to propagate these
changes through the SAN fabric.
In order to mitigate I/O disruption, it is recommended to instruct the HBA driver to wait up to 60 seconds
before marking a port as down or failed. For volumes with multiple access paths managed by multipathd, if
all paths become inaccessible, multipathd starts queuing I/O until one or more paths have recovered. This
allows the storage system sufficient time to relocate the WWPN identities of the failed ports to active ports
and propagate the changes through the SAN fabric.
The HBA PortDown timeout parameter dictates how long a Linux system waits before destroying a connection
after losing its connectivity with the port. The PortDown timeout is generally set to 60 seconds for the SC
Series storage system. Consult the Dell EMC Storage Compatibility Matrix for the most up-to-date
recommended values.
To configure the PortDown timeout in the HBA driver module, add or append the following options to the
appropriate HBA driver module configuration file (section 2.1) and reload the driver module (section 2.1.3).
options qla2xxx qlport_down_retry=60 #For QLogic
options lpfc lpfc_devloss_tmo=60 #For Emulex
2.1.3 Reloading modprobe
After updating the HBA driver module configuration file, reload the modprobe facility for the new or updated
configuration to take effect.