[ofa-general] OpenSM initialization error

[Yicheng Jia]

It occurs with Qlogic switch, the error is reported on the OpenSM log 
file. Both OpenSM and mthca driver are running on QNX environment, so 
there's no handy tool available. What else can I do?

Thanks!

Yicheng Jia





Hal Rosenstock <hal.rosenstock at gmail.com> 
03/26/2009 12:53 PM

To
Yicheng Jia <YJia at tmriusa.com>
cc
general at lists.openfabrics.org, hnrose at comcast.net
Subject
Re: [ofa-general] OpenSM initialization error






On Thu, Mar 26, 2009 at 1:46 PM, Yicheng Jia <YJia at tmriusa.com> wrote:
>
>> You shouldn't be running different flavor SMs on the same subnet.
> The OpenSM will go to standby if there's another SM exists.

Some SM will go to standby but that's not the issues with mixing SMs
are well beyond the simple election/handover protocol.

>> Are you sure there's no SM running here other than OpenSM ?
> Actually this error always happens on Qlogic unmanaged switch, which I
> believe no SM running on the subnet.

Do you mean on or with QLogic switch ?

What does saquery -s say ?

-- Hal

> It always takes more than 1 minute for the error disappear.

> Thanks!
>
> Yicheng Jia
>
>
>
>
> Hal Rosenstock <hal.rosenstock at gmail.com>
>
> 03/26/2009 12:13 PM
>
> To
> Yicheng Jia <YJia at tmriusa.com>
> cc
> general at lists.openfabrics.org, hnrose at comcast.net
> Subject
> Re: [ofa-general] OpenSM initialization error
>
>
>
>
> On Thu, Mar 26, 2009 at 12:01 PM, Yicheng Jia <YJia at tmriusa.com> wrote:
>>
>> LID 1 is the node that OpenSM runs on. Cisco SFS7000P is managed 
switch,
>> SM
>> is running all the time.
>
> You shouldn't be running different flavor SMs on the same subnet.
>
>> This occurs during the system starts up. At first Mthca driver 1.3 
starts
>> on
>> all nodes, after the driver is up, the OpenSM starts on one node, then 
the
>> error shows up. It also occurs on Qlogic unmanaged switch.
>
> Are you sure there's no SM running here other than OpenSM ?
>
> -- Hal
>
>> Is this related to hardware?
>
>> Thanks!
>>
>> Yicheng Jia
>>
>>
>>
>>
>> Hal Rosenstock <hal.rosenstock at gmail.com>
>>
>> 03/26/2009 10:32 AM
>>
>> To
>> Yicheng Jia <YJia at tmriusa.com>
>> cc
>> general at lists.openfabrics.org, hnrose at comcast.net
>> Subject
>> Re: [ofa-general] OpenSM initialization error
>>
>>
>>
>>
>> On Thu, Mar 26, 2009 at 11:16 AM, Yicheng Jia <YJia at tmriusa.com> wrote:
>>>
>>> Hi Hal,
>>>
>>>> What is your topology ?
>>>
>>> Several nodes directly connect to the switch.
>>>
>>>> Does this message persist or go away in the log ?
>>>
>>> It continues reporting this message until the subnet is up, then this
>>> message goes away. It lasts about 1 minute.
>>
>> What is LID 1 ? Is it the switch ? Is Cisco SM running ?
>>
>>>> Looks to me like you are using an OpenSM 3.1.11/OFED 1.3 or older.
>>>> Amongst other changes, this message has been downgraded from error to
>>>> debug in more recent versions.
>>>
>>> Do you know what could cause the problem? Is there any way to make the
>>> error
>>> time shorter?
>>
>> What's the exact scenario under which this occurs ?
>>
>> -- Hal
>>
>>> Thanks!
>>>
>>> Yicheng Jia
>>>
>>>
>>>
>>>
>>> Hal Rosenstock <hal.rosenstock at gmail.com>
>>>
>>> 03/26/2009 09:58 AM
>>>
>>> To
>>> Yicheng Jia <YJia at tmriusa.com>
>>> cc
>>> hnrose at comcast.net, general at lists.openfabrics.org
>>> Subject
>>> Re: [ofa-general] OpenSM initialization error
>>>
>>>
>>>
>>>
>>> 2009/3/25 Yicheng Jia <YJia at tmriusa.com>:
>>>>
>>>> Hello,
>>>>
>>>> I run into an initialization error during OpenSM start up. The error
>>>> continues for nearly 1 minutes before the subnet is up. In the log 
file,
>>>> it
>>>> reports that "__osm_ucast_mgr_process_port: ERR 3A08: No path to get 
to
>>>> LID
>>>> 1 from switch 0x66a00d90008bb".
>>>>
>>>> I am using Mellanox MHES18 HCA and Cisco SFS7000P switch. Do you know
>>>> what
>>>> the cause is and how to avoid it?
>>>
>>> What is your topology ?
>>>
>>> Does this message persist or go away in the log ?
>>>
>>> Looks to me like you are using an OpenSM 3.1.11/OFED 1.3 or older.
>>> Amongst other changes, this message has been downgraded from error to
>>> debug in more recent versions.
>>>
>>> -- Hal
>>>
>>>>
>>>> Thanks!
>>>>
>>>> Yicheng Jia
>>>>
>>>>
>>>>
>>>>
>>>>
>>>> 
_____________________________________________________________________________
>>>> Scanned by IBM Email Security Management Services powered by
>>>> MessageLabs.
>>>> For more information please visit http://www.ers.ibm.com
>>>>
>>>>
>>>>
>>>> 
_____________________________________________________________________________
>>>>
>>>> _______________________________________________
>>>> general mailing list
>>>> general at lists.openfabrics.org
>>>> http://lists.openfabrics.org/cgi-bin/mailman/listinfo/general
>>>>
>>>> To unsubscribe, please visit
>>>> http://openib.org/mailman/listinfo/openib-general
>>>>
>>>
>>>
>>>
>>> 
_____________________________________________________________________________
>>> Scanned by IBM Email Security Management Services powered by 
MessageLabs.
>>> For more information please visit http://www.ers.ibm.com
>>>
>>>
>>> 
_____________________________________________________________________________
>>>
>>>
>>>
>>>
>>> 
_____________________________________________________________________________
>>> Scanned by IBM Email Security Management Services powered by 
MessageLabs.
>>> For more information please visit http://www.ers.ibm.com
>>>
>>>
>>> 
_____________________________________________________________________________
>>>
>>
>>
>> 
_____________________________________________________________________________
>> Scanned by IBM Email Security Management Services powered by 
MessageLabs.
>> For more information please visit http://www.ers.ibm.com
>>
>> 
_____________________________________________________________________________
>>
>>
>>
>> 
_____________________________________________________________________________
>> Scanned by IBM Email Security Management Services powered by 
MessageLabs.
>> For more information please visit http://www.ers.ibm.com
>>
>> 
_____________________________________________________________________________
>>
>
> 
_____________________________________________________________________________
> Scanned by IBM Email Security Management Services powered by 
MessageLabs.
> For more information please visit http://www.ers.ibm.com
> 
_____________________________________________________________________________
>
>
> 
_____________________________________________________________________________
> Scanned by IBM Email Security Management Services powered by 
MessageLabs.
> For more information please visit http://www.ers.ibm.com
> 
_____________________________________________________________________________
>

_____________________________________________________________________________
Scanned by IBM Email Security Management Services powered by MessageLabs. 
For more information please visit http://www.ers.ibm.com
_____________________________________________________________________________



_____________________________________________________________________________
Scanned by IBM Email Security Management Services powered by MessageLabs. For more information please visit http://www.ers.ibm.com
_____________________________________________________________________________
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.openfabrics.org/pipermail/general/attachments/20090326/e5e49281/attachment.html>