[openib-general] Connectivity
Sean Hubbell
shubbell at dbresearch.net
Fri Sep 2 10:42:43 PDT 2005
Hal Rosenstock wrote:
>Hi again Sean,
>
>On Fri, 2005-09-02 at 13:13, Sean Hubbell wrote:
>
>Thanks for bearing with this process...
>
>
>
>>Sep 02 11:44:38 [447FF960] -> SA MAD dump:
>> base_ver................0x1
>> mgmt_class..............0x3
>> class_ver...............0x2
>> method..................0x2 (SubnAdmSet)
>> status..................0x0
>> resv....................0x0
>> trans_id................0x7333c794d
>> attr_id.................0x38
>>(MCMemberRecord)
>> resv1...................0x0
>> attr_mod................0x0
>> rmpp_version............0x0
>> rmpp_type...............0x0
>> rmpp_flags..............0x0
>> rmpp_status.............0x0
>> seg_num.................0x0
>> payload_len/new_win.....0x0
>> sm_key..................0x0000000000000000
>> attr_offset.............0x0
>> resv2...................0x0
>> comp_mask...............0x00000000000130c7
>>
>>
>>Sep 02 11:44:38 [447FF960] -> __osm_sa_mad_ctrl_process: [
>>Sep 02 11:44:38 [447FF960] -> __osm_sa_mad_ctrl_process: Posting
>>Dispatcher message OSM_MSG_MAD_MCMEMBER_RECORD.
>>Sep 02 11:44:38 [447FF960] -> __osm_sa_mad_ctrl_process: ]
>>Sep 02 11:44:38 [42FFF960] -> osm_mcmr_rcv_process: [
>>Sep 02 11:44:38 [42FFF960] -> osm_mcmr_rcv_join_mgrp: [
>>Sep 02 11:44:38 [447FF960] -> __osm_sa_mad_ctrl_rcv_callback: ]
>>Sep 02 11:44:38 [42FFF960] -> osm_mcmr_rcv_join_mgrp: Dump of incoming
>>record.
>>Sep 02 11:44:38 [42FFF960] -> MCMember Record dump:
>>
>>MGID....................0xff12401bffff0000 : 0x0000000000090a0b
>>
>>PortGid.................0xfe80000000000000 : 0x0005ad000003d221
>> qkey....................0xB1B
>> Mlid....................0x0
>> ScopeState..............0x1
>> Rate....................0x0
>> Mtu.....................0x0
>> TClass..................0x0
>> SLFlowLabelHopLimit.....0x0
>>
>>
>
>This looks like a create for 224.9.10.11. What was the response just
>below it ?
>
>
>
>>Attached is the results of grep 1307c osm.log > 1307c.log.
>>
>>
>
>Many of those are join failures but there are others like this:
>
>
>> comp_mask...............0x00000000000130c7
>> comp_mask...............0x00000000000130c7
>> comp_mask...............0x00000000000130c7
>> comp_mask...............0x00000000000130c7
>> comp_mask...............0x00000000000130c7
>> comp_mask...............0x00000000000130c7
>> comp_mask...............0x00000000000130c7
>> comp_mask...............0x00000000000130c7
>> comp_mask...............0x00000000000130c7
>>
>>
>
>I would like the part of the log around these if possible.
>
>Let me know you svn version and I will provide a patch to display the
>MGID when the join fails.
>
>Also, it may be getting time to use those multicast routing and tracing
>tools mentioned earlier.
>
>-- Hal
>
>
>
Results of ibroute:
ibroute -M 4 0xc010 0xc020
Multicast mlids [0xc010-0xc020] of switch Lid 0x4 guid
0x0002c9010c584d40 (MT47396 Infiniscale-III Mellanox Technologies):
0 1 2
Ports: 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4
MLid
0 valid mlids dumped
Results for ibtracert:
>ibtracert -m 0xc000 4 16
>warn: [5512] find_target_portguid: can't find to port
>
>ibtracert: error: can't find a multicast route from src to dest
>
>
>
More information about the general
mailing list