Hi guys
I have 2 IBM servers presenting this issue in two different locations I mean by this that these are even in different countries so Storage-San Switch and everything is not shared
In one of this server VDP bakcups were failing and the other one 2 VMs were having a lot IO wait CPU, data base connections timeouts as soon as I moved these VMs out of these two server and VMs and backups are working fine
this warning repeats like 100 times per hour in each server
2014-02-20T19:55:47.339Z cpu5:33372)WARNING: NMP: nmp_DeviceRequestFastDeviceProbe:237: NMP device "eui.001738000b650371" state in doubt; requested fast path state update...
2014-02-20T19:56:00.279Z cpu5:33372)WARNING: NMP: nmp_DeviceRequestFastDeviceProbe:237: NMP device "eui.001738000b6501d5" state in doubt; requested fast path state update...
2014-02-20T19:56:07.568Z cpu4:33372)WARNING: NMP: nmp_DeviceRequestFastDeviceProbe:237: NMP device "eui.001738000b650371" state in doubt; requested fast path state update...
2014-02-20T19:56:15.347Z cpu4:33372)WARNING: NMP: nmp_DeviceRequestFastDeviceProbe:237: NMP device "eui.001738000b650371" state in doubt; requested fast path state update...
2014-02-20T19:58:11.337Z cpu0:33372)WARNING: NMP: nmp_DeviceRequestFastDeviceProbe:237: NMP device "eui.001738000b650371" state in doubt; requested fast path state update...
by the way this is the HBA: QLogic Corp ISP2532-based 8Gb Fibre Channel to PCI Express HBA
but I could not find any hab under /proc/scsi there is no reference to qlogic or HBas
any idea what to check or update? or how to fix it
recent change was these servers were both updated to ESXi 5.5
thanks a lot
anyone guys?
This is interesting, I observed identical behaviour across all of my vSphere 5.0 hosts as still don't have a clear understanding of exactly what is causing it. In my case I see it on HP and Dell servers and on both Qlogic and Emulex HBA's. I'll forward on the details when I get into the office.
Cheers,
Jon
in 5.5 there is a concept of native drivers.. Seems your environment is using the same.
to view the similar details of proc scsi .. Look using "/usr/lib/vmware/vmkmgmt_keyval/vmkmgmt_keyval -a" which lists all keys.
for eg: just for listing devices through emulex adapter this is the command /usr/lib/vmware/vmkmgmt_keyval/vmkmgmt_keyval -i Emulex/lpfc
well I don't know about your VMs but in my case like I said is affecting some VMs performance which is not good
checking with your command
I found FW and Driver version
QLogic PCI to Fibre Channel Host Adapter for QMI3572FC Firmware version 5.09.00 (90d5), Driver version 1.0.12.0
Looks like FW is OK but driver won't match according to this so I am thinking to try updating driver version might help I don't know, well I see there is 1.1.20. too
VMware Compatibility Guide: I/O Device Search
ESXi 5.5 | qlnativefc version 1.1.7.0-1 | 5.09.00 | async |
Hi,
had identical problem. Updated drivers. Set portcfgfillword <Portnumber> 3 on brocadeswitch:
http://www-304.ibm.com/support/docview.wss?uid=isg3T1013082
Essentially, it boiled down to a defective SFP Module. Replaced it, now everything works fine.
Cheers Mike
Hi mrueppcemm ,
How to make sure the defective SFP Module? I can not find any CRC error in the interface of switch port, but can see many package droped in some interface.