[Lustre-discuss] Lustre-1.8.1.1 over o2ib gives Input/Output error while executing lctl ping

[Vipul Pandya]

Hi All,

 

I am trying to run Lustre over iWARP. For this I have compiled
Lustre-1.8.1.1 with linux-2.6.18-128.7.1 source and OFED-1.5 source.

I have installed all the required rpms for lustre.

 

After this I booted into  the lustre patched kernel and gave the
following option in /etc/modprobe.conf for lnet to work with o2ib

#> cat /etc/modprobe.conf

options lnet networks="o2ib0(eth2)"

 

I loaded our RDMA adapter modules and the lnet and ko2iblnd modules as
follows:

#> modprobe cxgb3

#> modprobe iw_cxgb3

#> modprobe rdma_ucm

#> modprobe lnet

#> modprobe ko2iblnd

 

I was able to load all the modules successfully.

 

Then I assigned the ip address to eth2 interface and brought it up

#> ifconfig eth2 102.88.88.188/24 up

#> ifconfig

eth0      Link encap:Ethernet  HWaddr 00:30:48:C7:8F:8E

          inet addr:10.193.184.188  Bcast:10.193.187.255
Mask:255.255.252.0

          inet6 addr: fe80::230:48ff:fec7:8f8e/64 Scope:Link

          UP BROADCAST RUNNING MULTICAST  MTU:1500  Metric:1

          RX packets:13224 errors:0 dropped:0 overruns:0 frame:0

          TX packets:797 errors:0 dropped:0 overruns:0 carrier:0

          collisions:0 txqueuelen:1000

          RX bytes:1523344 (1.4 MiB)  TX bytes:203205 (198.4 KiB)

          Memory:dea20000-dea40000

 

eth2      Link encap:Ethernet  HWaddr 00:07:43:05:07:35

          inet addr:102.88.88.188  Bcast:102.88.88.255
Mask:255.255.255.0

          inet6 addr: fe80::207:43ff:fe05:735/64 Scope:Link

          UP BROADCAST RUNNING MULTICAST  MTU:1500  Metric:1

          RX packets:153 errors:0 dropped:0 overruns:0 frame:0

          TX packets:47 errors:0 dropped:0 overruns:0 carrier:0

          collisions:0 txqueuelen:1000

          RX bytes:22537 (22.0 KiB)  TX bytes:8500 (8.3 KiB)

          Interrupt:185 Memory:de801000-de801fff

 

lo        Link encap:Local Loopback

          inet addr:127.0.0.1  Mask:255.0.0.0

          inet6 addr: ::1/128 Scope:Host

          UP LOOPBACK RUNNING  MTU:16436  Metric:1

          RX packets:1607 errors:0 dropped:0 overruns:0 frame:0

          TX packets:1607 errors:0 dropped:0 overruns:0 carrier:0

          collisions:0 txqueuelen:0

          RX bytes:3196948 (3.0 MiB)  TX bytes:3196948 (3.0 MiB)

 

After this I tried to bring the lnet network up as follows:

#> lctl network up

LNET configured

 

Above command gave me following error in dmesg

#> dmesg

Lustre: Listener bound to eth2:102.88.88.188:987:cxgb3_0

Lustre: Register global MR array, MR size: 0xffffffff, array size: 2

fmr_pool: Device cxgb3_0 does not support FMRs

LustreError: 4134:0:(o2iblnd.c:1393:kiblnd_create_fmr_pool()) Failed to
create FMR pool: -38

Lustre: Added LNI 102.88.88.188 at o2ib [8/64/0/0]

 

I repeat the same procedure on the other node of lustre and found the
same result.

Then I tried to do lctl ping between two nodes of lustre, which gave me
following error:

 

#> lctl ping 102.88.88.184 at o2ib

failed to ping 102.88.88.184 at o2ib: Input/output error

 

dmesg has shown following error:

#> dmesg

LustreError: 2453:0:(o2iblnd.c:801:kiblnd_create_conn()) Can't create
QP: -12, send_wr: 2056, recv_wr: 18

Lustre: 2453:0:(o2iblnd_cb.c:1953:kiblnd_peer_connect_failed()) Deleting
messages for 102.88.88.184 at o2ib: connection failed

 

I found one thread where it has given the patch to support FMR in o2ib.
But I don't think this patch is applicable for lustre-1.8.1.1.

http://lists.lustre.org/pipermail/lustre-discuss/2008-February/006502.ht
ml

 

Can anyone please guide me on this.

 

Thank you very much in advance.

Vipul

 

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.lustre.org/pipermail/lustre-discuss-lustre.org/attachments/20100212/8273072a/attachment.htm>