OpenSM: reporting traps 129, 130, 131

OpenSM: reporting traps 129, 130, 131

[Yevgeny Kliteynik]

Hi Sasha,

I noticed that OpenSM doesn't send InformInfo on traps 129/130/131.
This is what osm_trap_rcv.c is doing:

322: static void trap_rcv_process_request(IN osm_sm_t * sm,
323: 				     IN const osm_madw_t * p_madw)
... 

435: 		if (ib_notice_is_generic(p_ntci) &&
436: 		    (p_ntci->g_or_v.generic.trap_num == CL_HTON16(129) ||
437: 		     p_ntci->g_or_v.generic.trap_num == CL_HTON16(130) ||
438: 		     p_ntci->g_or_v.generic.trap_num == CL_HTON16(131))) {
439: 			/* If this is a trap 129, 130, or 131 - then this is a
440: 			 * trap signaling a change on a physical port.
441: 			 * Mark the physp_change_trap flag as TRUE.
442: 			 */
443: 			physp_change_trap = TRUE;
...

539:	/* If we reached here due to trap 129/130/131 - do not need to do
540:	   the notice report. Just goto exit. We know this is the case
541:	   if physp_change_trap is TRUE. */
542:	if (physp_change_trap == TRUE)
543:		goto Exit;


Any particular reason why there's no reporting of these traps?

-- Yevgeny


--
To unsubscribe from this list: send the line "unsubscribe linux-rdma" in
the body of a message to majordomo-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

Re: OpenSM: reporting traps 129, 130, 131

[Hal Rosenstock]

On Sun, Nov 8, 2009 at 9:37 AM, Yevgeny Kliteynik
<kliteyn-LDSdmyG8hGV8YrgS2mwiifqBs+8SCbDb@public.gmane.org> wrote:
> Hi Sasha,
>
> I noticed that OpenSM doesn't send InformInfo on traps 129/130/131.
> This is what osm_trap_rcv.c is doing:
>
> 322: static void trap_rcv_process_request(IN osm_sm_t * sm,
> 323:                                 IN const osm_madw_t * p_madw)
> ...
> 435:            if (ib_notice_is_generic(p_ntci) &&
> 436:                (p_ntci->g_or_v.generic.trap_num == CL_HTON16(129) ||
> 437:                 p_ntci->g_or_v.generic.trap_num == CL_HTON16(130) ||
> 438:                 p_ntci->g_or_v.generic.trap_num == CL_HTON16(131))) {
> 439:                    /* If this is a trap 129, 130, or 131 - then this is
> a
> 440:                     * trap signaling a change on a physical port.
> 441:                     * Mark the physp_change_trap flag as TRUE.
> 442:                     */
> 443:                    physp_change_trap = TRUE;
> ...
>
> 539:    /* If we reached here due to trap 129/130/131 - do not need to do
> 540:       the notice report. Just goto exit. We know this is the case
> 541:       if physp_change_trap is TRUE. */
> 542:    if (physp_change_trap == TRUE)
> 543:            goto Exit;
>
>
> Any particular reason why there's no reporting of these traps?

AFAIK it's been this way for at least the last 5 or 6 years.

A practical consideration in changing this is that these traps are the
ones for which babbling port was implemented since they do not obey
the trap rate so there's a large downside to adding the reports for
these. If that is to be done, then perhaps this should be done based
on some option.

-- Hal

>
> -- Yevgeny
>
>
> --
> To unsubscribe from this list: send the line "unsubscribe linux-rdma" in
> the body of a message to majordomo-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
> More majordomo info at  http://vger.kernel.org/majordomo-info.html
>
--
To unsubscribe from this list: send the line "unsubscribe linux-rdma" in
the body of a message to majordomo-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

Re: OpenSM: reporting traps 129, 130, 131

[Sasha Khapyorsky]

On 08:46 Thu 12 Nov     , Hal Rosenstock wrote:
> >
> > Any particular reason why there's no reporting of these traps?
> 
> AFAIK it's been this way for at least the last 5 or 6 years.

Yes, this is what I found too tracking down git/svn history.

> A practical consideration in changing this is that these traps are the
> ones for which babbling port was implemented since they do not obey
> the trap rate so there's a large downside to adding the reports for
> these. If that is to be done, then perhaps this should be done based
> on some option.

I don't think that we need new option for this, instead we should care
that OpenSM is not overloaded by such trap floods.

Sasha
--
To unsubscribe from this list: send the line "unsubscribe linux-rdma" in
the body of a message to majordomo-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

Re: OpenSM: reporting traps 129, 130, 131

[Hal Rosenstock]

On Thu, Nov 12, 2009 at 10:42 AM, Sasha Khapyorsky <sashak-smomgflXvOZWk0Htik3J/w@public.gmane.org> wrote:
> On 08:46 Thu 12 Nov     , Hal Rosenstock wrote:
>> >
>> > Any particular reason why there's no reporting of these traps?
>>
>> AFAIK it's been this way for at least the last 5 or 6 years.
>
> Yes, this is what I found too tracking down git/svn history.
>
>> A practical consideration in changing this is that these traps are the
>> ones for which babbling port was implemented since they do not obey
>> the trap rate so there's a large downside to adding the reports for
>> these. If that is to be done, then perhaps this should be done based
>> on some option.
>
> I don't think that we need new option for this, instead we should care
> that OpenSM is not overloaded by such trap floods.

We've cared for several years now (babbling port policy was introduced
into the master on 7/6/09) but this issue still persists. Are you
saying no reports for these traps until the trap rate is obeyed ?

-- Hal

>
> Sasha
>
--
To unsubscribe from this list: send the line "unsubscribe linux-rdma" in
the body of a message to majordomo-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

Re: OpenSM: reporting traps 129, 130, 131

[Hal Rosenstock]

On Thu, Nov 12, 2009 at 11:38 AM, Hal Rosenstock
<hal.rosenstock-Re5JQEeQqe8AvxtiuMwx3w@public.gmane.org> wrote:
> On Thu, Nov 12, 2009 at 10:42 AM, Sasha Khapyorsky <sashak@voltaire.com> wrote:
>> On 08:46 Thu 12 Nov     , Hal Rosenstock wrote:
>>> >
>>> > Any particular reason why there's no reporting of these traps?
>>>
>>> AFAIK it's been this way for at least the last 5 or 6 years.
>>
>> Yes, this is what I found too tracking down git/svn history.
>>
>>> A practical consideration in changing this is that these traps are the
>>> ones for which babbling port was implemented since they do not obey
>>> the trap rate so there's a large downside to adding the reports for
>>> these. If that is to be done, then perhaps this should be done based
>>> on some option.
>>
>> I don't think that we need new option for this, instead we should care
>> that OpenSM is not overloaded by such trap floods.
>
> We've cared for several years now (babbling port policy was introduced
> into the master on 7/6/09)

I meant 7/6/07

> but this issue still persists. Are you
> saying no reports for these traps until the trap rate is obeyed ?
>
> -- Hal
>
>>
>> Sasha
>>
>
--
To unsubscribe from this list: send the line "unsubscribe linux-rdma" in
the body of a message to majordomo-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

Re: OpenSM: reporting traps 129, 130, 131

[Sasha Khapyorsky]

On 11:38 Thu 12 Nov     , Hal Rosenstock wrote:
> 
> We've cared for several years now (babbling port policy was introduced
> into the master on 7/6/09) but this issue still persists. Are you
> saying no reports for these traps until the trap rate is obeyed ?

If existing trap rate filtering in OpenSM is sufficient we can add
reporting (if somebody needs this). I just don't think that we need to
introduce temporary options.

Sasha
--
To unsubscribe from this list: send the line "unsubscribe linux-rdma" in
the body of a message to majordomo-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

Re: OpenSM: reporting traps 129, 130, 131

[Hal Rosenstock]

On Thu, Nov 12, 2009 at 3:36 PM, Sasha Khapyorsky <sashak-smomgflXvOZWk0Htik3J/w@public.gmane.org> wrote:
> On 11:38 Thu 12 Nov     , Hal Rosenstock wrote:
>>
>> We've cared for several years now (babbling port policy was introduced
>> into the master on 7/6/07) but this issue still persists. Are you
>> saying no reports for these traps until the trap rate is obeyed ?
>
> If existing trap rate filtering in OpenSM is sufficient we can add
> reporting (if somebody needs this).

Right but the key is whether it is sufficient or not and trap
reporting can be expensive so I was thinking it could be based on the
existing babbling_port_policy being off so it's not a new option but
further use of an existing one.

> I just don't think that we need to introduce temporary options.

I think we're past the point of temporary on this.

-- Hal

>
> Sasha
>
--
To unsubscribe from this list: send the line "unsubscribe linux-rdma" in
the body of a message to majordomo-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html