[Users] HP BLc QLogic 4X QDR IB Switch oddness

[Hal Rosenstock]

Andrei,

On Fri, May 24, 2013 at 10:13 AM, Andrei Mikhailovsky <andrei at arhont.com>wrote:

> Hal,
>
> The trouble is i've tried swapping the cards and even the IB switch, makes
> no difference.
>
> The link state changes that i've mentioned before keeps cycling prior to
> being activated. as i've mentioned it takes various amount of time before
> they are being activated. However, once active they stay active for months.
> there are no errors or any other problems once the link becomes active.
>
> The trouble is to become active it takes a while, which is not right )))
>

The (physical) link state cycling is indicative of some physical
connectivity issue. If swapping cards doesn't work, I'm not sure what to
tell you. Maybe it's a backplane issue but likely you can't swap the
chassis.

-- Hal


>
> Andrei
> ------------------------------
> *From: *"Hal Rosenstock" <hal.rosenstock at gmail.com>
> *To: *"Andrei Mikhailovsky" <andrei at arhont.com>
> *Cc: *users at lists.openfabrics.org
> *Sent: *Friday, 24 May, 2013 2:47:48 PM
>
> *Subject: *Re: [Users] HP BLc QLogic 4X QDR IB Switch oddness
>
> Andrei,
>
> On Fri, May 24, 2013 at 9:45 AM, Andrei Mikhailovsky <andrei at arhont.com>wrote:
>
>> Hal,
>>
>> The physical state changes every few seconds and goes from DOWN/Polling
>> to DOWN/Disabled to DOWN/(something else) don't remember exactly. It just
>> keeps cycling through them. I've never seen the ports in INIT state, which
>> I think it should start with when there is a connectivity.
>>
>
> So there's some physical connectivity issue. The link is constantly trying
> to negotiate and come up to LinkUp but even when it gets there it doesn't
> stay there for long.
>
> I'd start by reseating the cards which are involved with this. Hopefully
> that will help to stabilize things.
>
> -- Hal
>
>
>>
>> Thanks
>> ------------------------------
>> *From: *"Hal Rosenstock" <hal.rosenstock at gmail.com>
>> *To: *"Andrei Mikhailovsky" <andrei at arhont.com>
>> *Cc: *users at lists.openfabrics.org
>> *Sent: *Friday, 24 May, 2013 2:11:24 PM
>> *Subject: *Re: [Users] HP BLc QLogic 4X QDR IB Switch oddness
>>
>>
>> Hi Andrei,
>>
>> On Fri, May 24, 2013 at 8:51 AM, Andrei Mikhailovsky <andrei at arhont.com>wrote:
>>
>>> Hello guys,
>>>
>>> I was wondering if anyone on the mailing list has experienced an odd
>>> behaviour with HP BLc QLogic 4X QDR IB Switch (505958-B21)? I've tried to
>>> get HP to solve this issue, but it resulted to HP basically saying that
>>> they do not have Infiniband specialists who could help me.
>>>
>>> My problem is that every time I reboot the blade servers which are
>>> connected to the above blade switch the switch ports take ages to Activate.
>>> The time it takes when the servers have:
>>>
>>> State: Active
>>> Physical state: LinkUp
>>>
>>> is random and varies from several minutes to over 10 hours. I couldn't
>>> find a correlation or any consistency between the servers or switch ports.
>>> The same server could connect within minutes after a reboot, but could take
>>> hours following the next reboot.
>>>
>>> However, the server which are connected to the same switch with a cable
>>> (not interconnected blade servers) do not have this issue and get port
>>> state Active in seconds every time.
>>>
>> What is the physical state of these ports when they're not active (on
>> both the switch and server side) ? Are they LinkUp or something else ?
>>
>>
>>>
>>> I have installed the latest available firmware and I have tried chaging
>>> blade mezz cards, IB switch and the blade enclosure, different versions of
>>> OFED / opensm. It stays the same.
>>>
>>> Has anyone come across this kind of behaviour?
>>>
>>
>> Have you tried reseating your (server) cards ?
>>
>> Also, would you comment on the topology ? How large a subnet is this ?
>>
>> -- Hal
>>
>>
>>>
>>> Thanks
>>>
>>> Andrei
>>>
>>>
>>> _______________________________________________
>>> Users mailing list
>>> Users at lists.openfabrics.org
>>> http://lists.openfabrics.org/cgi-bin/mailman/listinfo/users
>>>
>>>
>>
>>
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.openfabrics.org/pipermail/users/attachments/20130524/82ca9b87/attachment.html>