Mailing List Archive: MLX Capability

MLX Capability

Feb 19, 2014, 11:22 PM

Post #1 of 4 (1615 views)

Dear Team,
We have 2 Brocade MLX-16s at the distribution layer of our DC
connecting to FCX as TOR switches. We were thinking to upgrade the DC
capacity where the servers connect to 10G links instead of current 1G. Thus
replace the FCXs with something newer that is 10G capable, may be a VDX.
However, on the MLX we only have 8x10G and 4x10G cards. I was thinking if
these can be replaced with 24x10G cards without an issue. Do we need to
upgrade the SFMs to HSF to achieve this? And also, is there a possibility
of having 2x40G LP on the MLX-16? Please note that we have MLX-16s not
MLX-e routers.

BR
Salman

Re: MLX Capability [ In reply to ]

martijnschmidt at i3d

Feb 20, 2014, 2:50 AM

Post #2 of 4 (1565 views)

Permalink

Hi Salman,

You need a MLXe backplane to install some of the newer blades, amongst others the 24x10G, 40G and 100G linecards.

Please note hSFM are already required to use 8x10G blades in a MLX chassis, you can only run gen1 cards such as the 4x10G blade with "old/regular" SFM.

On 20 February 2014 08:22:22 CET, salman sadiq <salmanravian@gmail.com> wrote:
>Dear Team,
> We have 2 Brocade MLX-16s at the distribution layer of our DC
>connecting to FCX as TOR switches. We were thinking to upgrade the DC
>capacity where the servers connect to 10G links instead of current 1G.
>Thus
>replace the FCXs with something newer that is 10G capable, may be a
>VDX.
>However, on the MLX we only have 8x10G and 4x10G cards. I was thinking
>if
>these can be replaced with 24x10G cards without an issue. Do we need to
>upgrade the SFMs to HSF to achieve this? And also, is there a
>possibility
>of having 2x40G LP on the MLX-16? Please note that we have MLX-16s not
>MLX-e routers.
>
>BR
>Salman
>
>
>------------------------------------------------------------------------
>
>_______________________________________________
>foundry-nsp mailing list
>foundry-nsp@puck.nether.net
>http://puck.nether.net/mailman/listinfo/foundry-nsp

--
Sent from my Android phone with K-9 Mail. Please excuse my brevity.

Re: MLX Capability [ In reply to ]

salmanravian at gmail

Feb 20, 2014, 3:10 AM

Post #3 of 4 (1566 views)

Permalink

Thank you for the information. You are right about the 8x10G cards
requiring hSFM. In fact we already have hSFMs installed. For the 24x10G and
40G modules, will they simply not work in an MLX chassis? Or is the
limitation coming from the chassis t'put? In case it is a t'put limitation
we can calculate the right mix of 10G 24 port and 4/8 port LPs. Of course
not all the 16 slots are going to contain 24x10G cards.

Thank you.

On Thu, Feb 20, 2014 at 2:50 PM, i3D.net - Martijn Schmidt <
martijnschmidt@i3d.net> wrote:

> Hi Salman,
>
> You need a MLXe backplane to install some of the newer blades, amongst
> others the 24x10G, 40G and 100G linecards.
>
> Please note hSFM are already required to use 8x10G blades in a MLX
> chassis, you can only run gen1 cards such as the 4x10G blade with
> "old/regular" SFM.
>
> On 20 February 2014 08:22:22 CET, salman sadiq <salmanravian@gmail.com>
> wrote:
>
>> Dear Team,
>> We have 2 Brocade MLX-16s at the distribution layer of our
>> DC connecting to FCX as TOR switches. We were thinking to upgrade the DC
>> capacity where the servers connect to 10G links instead of current 1G. Thus
>> replace the FCXs with something newer that is 10G capable, may be a VDX.
>> However, on the MLX we only have 8x10G and 4x10G cards. I was thinking if
>> these can be replaced with 24x10G cards without an issue. Do we need to
>> upgrade the SFMs to HSF to achieve this? And also, is there a possibility
>> of having 2x40G LP on the MLX-16? Please note that we have MLX-16s not
>> MLX-e routers.
>>
>> BR
>> Salman
>>
>> ------------------------------
>>
>> foundry-nsp mailing list
>> foundry-nsp@puck.nether.net
>> http://puck.nether.net/mailman/listinfo/foundry-nsp
>>
>>
> --
> Sent from my Android phone with K-9 Mail. Please excuse my brevity.
>

Re: MLX Capability [ In reply to ]

martijnschmidt at i3d

Feb 20, 2014, 3:14 AM

Post #4 of 4 (1561 views)

Permalink

Hi Salman,

I never put it to the test, but these cards will probably just not function in a MLX at all.

On 20 February 2014 12:10:34 CET, salman sadiq <salmanravian@gmail.com> wrote:
>Thank you for the information. You are right about the 8x10G cards
>requiring hSFM. In fact we already have hSFMs installed. For the 24x10G
>and
>40G modules, will they simply not work in an MLX chassis? Or is the
>limitation coming from the chassis t'put? In case it is a t'put
>limitation
>we can calculate the right mix of 10G 24 port and 4/8 port LPs. Of
>course
>not all the 16 slots are going to contain 24x10G cards.
>
>Thank you.
>
>
>On Thu, Feb 20, 2014 at 2:50 PM, i3D.net - Martijn Schmidt <
>martijnschmidt@i3d.net> wrote:
>
>> Hi Salman,
>>
>> You need a MLXe backplane to install some of the newer blades,
>amongst
>> others the 24x10G, 40G and 100G linecards.
>>
>> Please note hSFM are already required to use 8x10G blades in a MLX
>> chassis, you can only run gen1 cards such as the 4x10G blade with
>> "old/regular" SFM.
>>
>> On 20 February 2014 08:22:22 CET, salman sadiq
><salmanravian@gmail.com>
>> wrote:
>>
>>> Dear Team,
>>> We have 2 Brocade MLX-16s at the distribution layer of
>our
>>> DC connecting to FCX as TOR switches. We were thinking to upgrade
>the DC
>>> capacity where the servers connect to 10G links instead of current
>1G. Thus
>>> replace the FCXs with something newer that is 10G capable, may be a
>VDX.
>>> However, on the MLX we only have 8x10G and 4x10G cards. I was
>thinking if
>>> these can be replaced with 24x10G cards without an issue. Do we need
>to
>>> upgrade the SFMs to HSF to achieve this? And also, is there a
>possibility
>>> of having 2x40G LP on the MLX-16? Please note that we have MLX-16s
>not
>>> MLX-e routers.
>>>
>>> BR
>>> Salman
>>>
>>> ------------------------------
>>>
>>> foundry-nsp mailing list
>>> foundry-nsp@puck.nether.net
>>> http://puck.nether.net/mailman/listinfo/foundry-nsp
>>>
>>>
>> --
>> Sent from my Android phone with K-9 Mail. Please excuse my brevity.
>>

--
Sent from my Android phone with K-9 Mail. Please excuse my brevity.