On Thu, 2013-07-18 at 15:35 +0000, Love, Robert W wrote:
> > Sent: Thursday, July 18, 2013 3:18 AM
> > To: [email protected]
> > Subject: [Open-FCoE] stuck with vn2vn
> >
> >
> > Good day,
> > I trying to test sw target and initiator.
> > My config: two servers with intel x540 connected back-to-back(vn2vn).
> > Centos 6.4, software from official centos repo, not latest!:
> > fcoe-target-utils.noarch 2.0rc1.fb16-3.el6 @base
> > fcoe-utils.x86_64 1.0.24-2.el6 @base
> > node0 target
> > node1 initiator
> > What I've done:
> > 1) echo eth1 > /sys/module/libfcoe/parameters/create_vn2vn
> > on both nodes
> > 2) on target
> > o- / .....................................................................
> > [...]
> > o- backstores ..........................................................
> > [...]
> > | o- block ................................................ [0 Storage
> > | Object]
> > | o- fileio ............................................... [1 Storage
> > | Object]
> > | | o- my_disk ................... [/tmp/os.img (1.0GiB) write-back
> > | | activated]
> > | o- pscsi ................................................ [0 Storage
> > | Object]
> > o- loopback ...................................................... [0
> > Targets]
> > o- tcm_fc ........................................................ [1
> > Targets]
> > o- 20:00:00:25:90:6d:43:3f ....................................... [enabled]
> > o- acls .......................................................... [1 ACL]
> > | o- 20:00:00:25:90:9c:63:ed .............................. [1 Mapped
> > | LUN]
> > | o- mapped_lun0 ............................ [lun0 fileio/my_disk (rw)]
> > o- luns .......................................................... [1 LUN]
> > o- lun0 ................................. [fileio/my_disk (/tmp/os.img)]
> > ----------------------------------------------
> > @node0 ~]# fcoeadm -i
> > Description: Ethernet Controller 10-Gigabit X540-AT2
> > Revision: 01
> > Manufacturer: Intel Corporation
> > Serial Number: Unknown
> > Driver: ixgbe 3.9.15-k
> > Number of Ports: 1
> > Symbolic Name: fcoe v0.1 over eth1
> > OS Device Name: host9
> > Node Name: 0x10000025906D433F
> > Port Name: 0x20000025906D433F
> > FabricName: 0x0000000000000000
> > Speed: 10 Gbit
> > Supported Speed: 1 Gbit, 10 Gbit
> > MaxFrameSize: 2112
> > FC-ID (Port ID): 0x00433F
> > State: Online
> > ----------------------------------------------
> > @node1 ~]# fcoeadm -i
> > Description: Ethernet Controller 10-Gigabit X540-AT2
> > Revision: 01
> > Manufacturer: Intel Corporation
> > Serial Number: Unknown
> > Driver: ixgbe 3.9.15-k
> > Number of Ports: 1
> > Symbolic Name: fcoe v0.1 over eth1
> > OS Device Name: host11
> > Node Name: 0x10000025909C63ED
> > Port Name: 0x20000025909C63ED
> > FabricName: 0x0000000000000000
> > Speed: 10 Gbit
> > Supported Speed: 1 Gbit, 10 Gbit
> > MaxFrameSize: 2112
> > FC-ID (Port ID): 0x0063ED
> > State: Online
> > ----------------------------------------------
> > @node1 ~]# fcping -h eth1 -F 0x00433F -c 3 Maximum ECHO data allowed by
> > the Fabric (0xfffffd) : 2108 bytes.
> > Maximum ECHO data allowed by the Source (0x0063ed) : 2108 bytes.
> > Maximum ECHO data allowed by the Target (0x00433f) : 2108 bytes.
> > Maximum ECHO data requested from user input (-s) : 32 (default 32) bytes.
> > Actual FC ELS ECHO data size used : 32 bytes.
> > Actual FC ELS ECHO payload size used : 36 bytes (including 4 bytes ECHO
> > command).
> > Sending FC ELS ECHO from 0x63ED (fc_host11) to 0x433F:
> > echo 1 accepted 0.290 ms
> > echo 2 accepted 0.320 ms
> > echo 3 accepted 0.402 ms
> > 3 frames sent, 3 received 0 errors, 0.000% loss, avg. rt time 0.337 ms
> > ----------------------------------------------
> > @node0 ~]# fcping -h eth1 -F 0x0063ED -c 3 Maximum ECHO data allowed by
> > the Fabric (0xfffffd) : 2108 bytes.
> > Maximum ECHO data allowed by the Source (0x00433f) : 2108 bytes.
> > Maximum ECHO data allowed by the Target (0x0063ed) : 2108 bytes.
> > Maximum ECHO data requested from user input (-s) : 32 (default 32) bytes.
> > Actual FC ELS ECHO data size used : 32 bytes.
> > Actual FC ELS ECHO payload size used : 36 bytes (including 4 bytes ECHO
> > command).
> > Sending FC ELS ECHO from 0x433F (fc_host9) to 0x63ED:
> > echo 1 accepted 0.332 ms
> > echo 2 accepted 0.403 ms
> > echo 3 accepted 0.377 ms
> > 3 frames sent, 3 received 0 errors, 0.000% loss, avg. rt time 0.371 ms
> > ----------------------------------------------
> > @node1 ~]# fcoeadm -t
> > Interface: eth1
> > Roles: FCP Target, FCP Initiator
> > Node Name: 0x10000025906D433F
> > Port Name: 0x20000025906D433F
> > Target ID: 0
> > MaxFrameSize: 2112
> > OS Device Name: rport-11:0-0
> > FC-ID (Port ID): 0x00433F
> > State: Online
> > LUN ID Device Name Capacity Block Size Description
> > ------ ----------- ---------- ---------- ----------------------------
> > 0 /dev/sdc 2586.08 PiB 1082720255 ........ . (rev )
> > ----------------------------------------------
> > Now I trying to fdisk /dev/sdc
> > But got:
> > "Unable to read /dev/sdc"
> > What I missing? Maybe I need to try latest versions of fcoe-target-utils and
> > fcoe-utils? latest kernel, another distro?
>
> Nothing jumps out at me from the output above. I don't know much about CentOS,
> but assuming that it's the same as RHEL6.4, I would think it should work.
>
> I don't have an answer to your problem. I'm happy to help debug. Another
> alternative
> would be to switch to Fedora and get things running there. There are a bunch
> of
> Fedora/RH guys on this list that I'm sure would be more eager to solve any
> Fedora issues.
>
Yes that is a options and I'm using FC17 over same ixgbe and that works
fine for me. My target cli config looks slightly different than your's
but that should not be issue here as you were able to see LUNs with
capacity info in fcoeadm, mean read capacity and inquiry must have
worked in your case and then not sure why fdisk wouldn't work
later...may be some permission related issue while selinux enabled, just
a hunch and possibly disabling seliunux may help here.
Also does simple 'dd' works in case just fdisk specific ?
Here is my FC17 working target config for your reference:
o- tcm_fc ............................. [2 Targets]
o- 20:00:00:18:8b:f7:2a:2f ...................... [enabled]
| o- acls ........................................... [1 ACL]
| | o- 20:00:00:00:00:01:02:14 ................ [2 Mapped LUNs]
| | o- mapped_lun1 ...............[lun1 fileio/lun1 (rw)]
| | o- mapped_lun2 .............. [lun2 fileio/lun2 (rw)]
| o- luns ................................. [2 LUNs]
| o- lun1 ....................... [fileio/lun1 (/luns/lun1.img)]
| o- lun2 ....................[fileio/lun2 (/luns/lun2.img)]
> Another point is that fcoe-utils is all just control plane stuff, it seems
> that your problem
> is with I/O, which makes me think it's a kernel problem. fcping is certainly
> in fcoe-utils,
> but it just passes commands to the kernel. fdisk will be generating reads from
> the kernel. I wouldn't think that an fcoe-utils upgrade would help.
> You could try turning on debug_logging to see if there's anything that pops
> up in the
> logs.
>
> echo 0xFF > /sys/module/libfc/parameters/debug_logging
>
> then run your fdisk again.
>
Also wireshark trace would be helpful beside looking at host and target
side debug logs with more debugging enabled.
Good mail post Alexy after doing setup config first and some debugging
using fcping etc.
//Vasu
_______________________________________________
fcoe-devel mailing list
[email protected]
http://lists.open-fcoe.org/mailman/listinfo/fcoe-devel
