On 4/28/2013 8:49 AM, Johnny Ström wrote:
Hello We have an Emulex LPe1150 that is causing FLOGI failure when the driver is loaded. We have updated the firmware on the card. It is working ok with an old HP MSA. But with an new MD36xxf so dose it not work. when the lpfc driver is loded so dose it take about 7 seconds to load and efter that so do we only see these LUN's: We do not see the other disk LUN's [16:0:0:0] disk DELL MD36xxf 0784 - [16:0:0:31] disk DELL Universal Xport 0784 - dmesg shows this after the driver is loaded. [15313.555729] scsi12 : Emulex LPe1150 1Gb PCIe Fibre Channel Adapter on PCI bus 0e device 00 irq 30 [15317.795558] lpfc 0000:0e:00.0: 1:1303 Link Up Event x1 received Data: x1 x1 x10 x2 x0 x0 0 [15317.795680] lpfc 0000:0e:00.0: 1:1309 Link Up Event npiv not supported in loop topology [15317.796111] lpfc 0000:0e:00.0: 1:(0):2858 FLOGI failure Status:x3/x18 TMO:x0 [15317.796423] lpfc 0000:0e:00.0: 1:(0):2858 FLOGI failure Status:x3/x18 TMO:x0 [15317.796726] lpfc 0000:0e:00.0: 1:(0):2858 FLOGI failure Status:x3/x18 TMO:x0 [15317.796825] lpfc 0000:0e:00.0: 1:(0):0100 FLOGI failure Status:x3/x18 TMO:x0 Could it be an bug in the lpfc driver when it is used with MD36xxf? Any suggestions what to do or try?
First thing to note here is that FLOGI is issued to log into a switch and not to an end-port. Next thing is that the logs you emailed show you are most probably in a "loop topology" and not in a Fabric topology. So, FLOGI messages are probably not relevant, unless you can confirm that LPe1150 is indeed connected to a FC switch.
Now, moving to the array side as a potential cause, disk arrays normally have setting associated with the front-side ports, which when not configured properly may cause issues with FC discovery (especially, when in loop topology). I certainly don't know if that is the case here, but I'd make sure to go thru the Dell MD36xx guide to ensure you have configured they array properly for loop topology. For example, some initiators won't "connect" with a array if the array port doesn't have Hard-addressing enabled when in loop topology.
One more thing (this is an obvious one, so sorry, if you have already done this one) -- I am assuming you have properly presented the disk LUNs in MD36xx to the LPe1105 port WWN. Right ?
Hope this helps, Sunny -- To unsubscribe from this list: send the line "unsubscribe linux-scsi" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html