08-26-2016 11:16 AM
help I need a coomand line for view the traffic hba en linux server Red Hat Enterprise Linux Server 7.0 (Maipo)
[root@srvbackup bin]# lspci | grep -i fibre
08:00.0 Fibre Channel: Brocade Communications Systems, Inc. 425/825/42B/82B 4Gbps/8Gbps PCIe dual port FC HBA (rev 01)
08:00.1 Fibre Channel: Brocade Communications Systems, Inc. 425/825/42B/82B 4Gbps/8Gbps PCIe dual port FC HBA (rev 01)
0b:00.0 Fibre Channel: Brocade Communications Systems, Inc. 425/825/42B/82B 4Gbps/8Gbps PCIe dual port FC HBA (rev 01)
0b:00.1 Fibre Channel: Brocade Communications Systems, Inc. 425/825/42B/82B 4Gbps/8Gbps PCIe dual port FC HBA (rev 01)
I need run backup and view the trafric in hba port.
thanks
08-26-2016 11:56 AM
08-28-2016 04:14 AM
As I understood you'd like to run a backup and check how it utilizes HBA port. Neither Linux nor NetBackup has such commands/tools. You need to understand what targets your HBA port works with and analyze your traffic to that targets (Or you can follow RedHat's recommendations https://access.redhat.com/solutions/440743). May be some HBA vendors provide analyzing tools for their HBAs but I don't remenber them.
P.S. Hope you didn't mean that you'd like to sniff FC traffic ;)
08-28-2016 05:34 AM
08-29-2016 05:22 AM
But he still needs to remember which target is available via which HBA-port and calculate total port utilization. Linux has iostat too :)
08-29-2016 06:35 AM
Best to monitor hba traffic from the switch side.
I have seen how a colleage logged into a brocade switch and issued commands to monitor traffic on one or more ports - server hba port and/or device port.
I doubt that there are OS-level commands for this.
If there are OS-level commands, the question will best be answered in a Linux forum.
08-29-2016 01:06 PM
Yes, you're absolutly right. Brocade has Fabric Watch utility and here are no pure OS-level commands neither in Solaris nor in Linux But if they had them it would be useless because admin must be familiar with SAN-topology after port. He should know switches' interconnect (ISL or truncks)bandwith and load and so on. Without whole picture it's just a numbers like "75% utilization" and you still need to go deeper into the darkness (storage array LUN performance, volume read parameters, OS I/O tuning vales, bus performance and so on and so om ) :)
08-30-2016 12:53 AM
Seems it's just you and me left in this conversation!
@Hernan_Peralta has not been back since he posted here....
09-02-2016 09:15 AM
4 fibers change master and the library and the drive no longer fall tanks
09-02-2016 11:15 AM