[PROXMOX 5.1] [lpfc] HBA Emulex LPe12000 error

[Oleksii Pechonkin]

Hi All

Proxmox 5.0 iso install - HBA see disk

Proxmox 5.1 iso install - HBA take error lpfc 0000:05:00.1: 10):0266 Issue NameServer Req x117 err 1 Data: x80000 x0

If Proxmox 5.0 (or clear debian 9) udgrade from pve-no-subscription - take error lpfc 0000:05:00.1: 10):0266 Issue NameServer Req x117 err 1 Data: x80000 x0

Yes i know don`t use pve-no-subscription in production but in 5.0 it's work. Please, who broke it, return it as it was .

Thank you. Have a nice day.

 

[borjamf]

Hi,

I got the same problem. I recently received two cabinets StorageTek, but I could not make it work in our testing node Proxmox 5.1, it shows the same error "Issue Nameserver....".

After I read this thread, I've installed Proxmox v5.0 in our testing node and now IT WORKS! I think something is broken in Proxmox v5.1, so hope anyone make a fix for this.

 

[Oleksii Pechonkin]

Hi All

4.13.13-5-pve still contains the same error

[ 4.890655] lpfc 0000:05:00.0: 00):0266 Issue NameServer Req x117 err 1 Data: x80000 x0

modinfo lpfc
filename: /lib/modules/4.13.13-5-pve/kernel/drivers/scsi/lpfc/lpfc.ko
version: 0:11.4.0.1
author: Broadcom
description: Emulex LightPulse Fibre Channel SCSI driver 11.4.0.1

 

[komspr]

Hi. I can confirm this issue. We have this on our servers.

description: Fibre Channel
product: Saturn-X: LightPulse Fibre Channel Host Adapter
vendor: Emulex Corporation

[ 4.015493] lpfc 0000:04:00.0: 0:1303 Link Up Event x1 received Data: x1 x1 x10 x2 x0 x0 0
[ 4.015570] lpfc 0000:04:00.0: 0:1309 Link Up Event npiv not supported in loop topology
[ 4.016671] lpfc 0000:04:00.0: 00):2858 FLOGI failure Status:x3/x18 TMO:x0 Data x1000 x0
[ 4.017627] lpfc 0000:04:00.0: 00):2858 FLOGI failure Status:x3/x18 TMO:x0 Data x1000 x0
[ 4.018561] lpfc 0000:04:00.0: 00):2858 FLOGI failure Status:x3/x18 TMO:x0 Data x1000 x0
[ 4.018636] lpfc 0000:04:00.0: 00):0100 FLOGI failure Status:x3/x18 TMO:x0
[ 4.019713] lpfc 0000:04:00.0: 00):0266 Issue NameServer Req x117 err 1 Data: x80000 x0

As we have Proxmox cluster we stuck with upgrading from 4.4 to 5.1.
Don't know if this is because of lpfc.ko module or some kernel issue. But Proxmox 5.0 had 4.10 kernel, 11.2.0.2 lpfc and Proxmox 5.1 has 4.13 kernel, 11.4.0.1 lpfc.
I tried to use lpfc.ko 11.0.0.10 which comes with 4.4 kernel, but no luck. Anyway we need update to resolve this issue.

Thank you.

 

[mir]

Hi. I can confirm this issue. We have this on our servers.

description: Fibre Channel
product: Saturn-X: LightPulse Fibre Channel Host Adapter
vendor: Emulex Corporation

[ 4.015493] lpfc 0000:04:00.0: 0:1303 Link Up Event x1 received Data: x1 x1 x10 x2 x0 x0 0
[ 4.015570] lpfc 0000:04:00.0: 0:1309 Link Up Event npiv not supported in loop topology
[ 4.016671] lpfc 0000:04:00.0: 00):2858 FLOGI failure Status:x3/x18 TMO:x0 Data x1000 x0
[ 4.017627] lpfc 0000:04:00.0: 00):2858 FLOGI failure Status:x3/x18 TMO:x0 Data x1000 x0
[ 4.018561] lpfc 0000:04:00.0: 00):2858 FLOGI failure Status:x3/x18 TMO:x0 Data x1000 x0
[ 4.018636] lpfc 0000:04:00.0: 00):0100 FLOGI failure Status:x3/x18 TMO:x0
[ 4.019713] lpfc 0000:04:00.0: 00):0266 Issue NameServer Req x117 err 1 Data: x80000 x0

As we have Proxmox cluster we stuck with upgrading from 4.4 to 5.1.
Don't know if this is because of lpfc.ko module or some kernel issue. But Proxmox 5.0 had 4.10 kernel, 11.2.0.2 lpfc and Proxmox 5.1 has 4.13 kernel, 11.4.0.1 lpfc.
I tried to use lpfc.ko 11.0.0.10 which comes with 4.4 kernel, but no luck. Anyway we need update to resolve this issue.

Thank you.

Have you tried consulting your hardware supplier?

 

[komspr]

No.

I have made a little investigation.

Ubuntu 16.04.3 LTS, kernel 4.10.0-28, lpfc 11.2.0.2 - no issue.
Ubuntu 17.10.1, kernel 4.13.0-21, lpfc 11.4.0.1 - issue.
Debian 9.3, kernel 4.9.0-4, lpfc 11.2.0.0 - no issue.
Fedora 27, kernel 4.13.9-300, lpfc 11.4.0.1 - issue.
Proxmox 5.0, kernel 4.10.15-1, lpfc 11.2.0.2 - no issue (thanks to Oleksii Pechonkin https://forum.proxmox.com/threads/proxmox-5-1-lpfc-hba-emulex-lpe12000-error.39179/).
Proxmox 5.1, kernel 4.13.13-5, lpfc 11.4.0.1 - issue.

So I think it's either kernel 4.13, or lpfc 11.4 problem...

 

[komspr]

Updated adapter firmware to latest - nothing changed, same issue.

 

[mamatov]

Hi,

I got the same problem. I have an old IBM DS3400 with two controllers.
I could not make it work in our testing node Proxmox 5.1 on "direct attach", it shows the same error "Issue Nameserver....".

If I installed proxmox 4.4, I can see my LUN.

With Proxmox 5.1, I found that if I add a fiber switch between the server and the DS3400 it is working as expected. I can see my LUN.

Is it working for you with a switch and not direct attach ?

 

[komspr]

Hi.
Unfortunately, I don't have a switch to test this workaround. And yes, I have this problem when using direct attach.

 

[mamatov]

Ah ok...

Anybody know why is it working with a switch and not on direct attach ?

 

[mir]

My guess is a missing update of the pciid database.

 

[komspr]

Updated to latest pci.ids - nothing changed.

 

[Mitterhuemer]

Hello,
we evaluate Proxmox with a HPE MSA2040 SAN, so we have no Subscription.

We also use the Emulex Fibre Channel Card but it does not work.

lpfc 0000:22:00.0:1: 1:1305 Link Down Event x4 received Data: x4 x20 x8011 x0 x0
lpfc 0000:22:00.0:1: 1:1303 Link Up Event x5 received Data: x5 x0 x20 x0 x0 x0 0
...
...
...

I have the same Problem with Proxmox 5.0

 

[mir]

Updated to latest pci.ids - nothing changed.

I was referring to the drivers pciid database;-)

 

[komspr]

Broadcom support wrote:
"Thanks for the logs, it sounds a valid bug, we will work a fix for it.
As your OS are not major Linux distributions, e.g. Ubuntu, there is only inbox driver support from OS vendor, and inbox driver is old, and the issue is possibly a known issue fixed on later version, but I have no newer driver for these OSes for verification.
If possible, could you please test our latest driver published on our web site on major OSes, such as SLES, RHEL and CentOS?
Driver download link:
https://www.broadcom.com/support/do...Pe12002+FC+Host+Bus+Adapter&po=&pa=Driver&dk="

I tried to adapt source code to Ubuntu kernel and compile it, but I had no success. I opened a ticket on Launchpad, so you may join and confirm this bug: https://bugs.launchpad.net/ubuntu/+source/linux/+bug/1746970

 

[mamatov]

Hi Komspr,

I confirm the bug on https://bugs.launchpad.net/ubuntu/+source/linux/+bug/1746970

 

[komspr]

Hi Komspr,

I confirm the bug on https://bugs.launchpad.net/ubuntu/+source/linux/+bug/1746970

Thank you. Don't forget to press "Does this bug affect you?" button to increase bug heat.

 

[oliwel]

In the launchpad item there is a solution: lpfc 11.4.0.4 which is in kernel 4.15 works.
Anybody knows if the pve kernel patches/sources are available so I can fix the module myself?

 

[fabian]

In the launchpad item there is a solution: lpfc 11.4.0.4 which is in kernel 4.15 works.
Anybody knows if the pve kernel patches/sources are available so I can fix the module myself?

the kernel sources are available (like all the others) via git.proxmox.com . I suggest checking mainline 4.14 and 4.13 next, depending on the outcome we can then further narrow down when the issue was introduced.

 

[oliwel]

Hello Fabian,

I did some more research, the last working kernel from the Ubuntu Xenial repository is
linux-image-4.10.0-42-generic_4.10.0-42.46~16.04.1_amd64.deb

The next available image linux-image-4.11.0-14-generic_4.11.0-14.20~16.04.1_amd64.deb is not working, I also tried several Ubuntu 4.13 and mainline kernels, none of them is working.

The first working kernel I found is mainline 4.14.1.

I think this commit is the fix for the problem github.com/torvalds/linux/commit/2877cbffb79ed121a6bcc5edbe629d3aba36cd29

Unfortunately the scsi subsystem was heavily refactored from 4.10 to 4.14 so I am unable to isolate a single patch to fix/reproduce the problem.

Are there any plans to upgrade to a newer kernel soon? I am still evaluating the PVE on my lab but need to bring those boxes in production within the next two weeks, so I need to decide weather to go with 5.0, try to patch myself or move to another solution...

Oliver