Re: [c-nsp] switch with 2x 10GBASE-T interfaces

2011-10-02 Thread Nikolay Shopik

Martin,

Yes, you right, my mistake, we are using C3KX-NM-10G, and at time when 
we bough it there no C3KX-NM-10GT modules, so I automatically think them.


You probably won't able to do so, because we even had problems with size 
of regular SFP(not rj45) on C3KX-NM-10G, so we had to replace them from 
different company. And from what I see sfp on 3750X-24S-S sit to each 
other very closely same as on C3KX-NM-10G.


Can't say anything about 4900 series.

On 02.10.2011 5:33, Martin T wrote:

Nikolay,
I'm afraid you are confusing this with C3KX-NM-10G module. Check the
table 3. C3KX-NM-10GT module is strictly 2x 10GBASE-T RJ45 ports(image
of the C3KX-NM-10GT: http://mcaf.ee/0bjty).

In general, WS-C3750X-24S-S + C3KX-NM-10GT seems to be a perfect
solution. WS-C3750X-24S-S has an IP Base image, which is enough
because I don't need advanced L3 features like EIGRP, OSPF, BGP and
IPv6 routing provided by IP Services. WS-C3750X-24S-S is easily
upgradable thanks to StackWise Plus and StackPower technologies.
Supports dual redundant power supplies and in overall is a nice 1U
feature-rich switch according to specifications.

Just one hesitation with WS-C3750X-24S-S..is it possible to insert
1000BASE-T SFP's to first 22 ports(I would use last two SFP ports for
1000BASE-LX10 SFP's)? I mean 1000BASE-T SFP's have RJ45 connector
which is bit bulky and I'm not sure if it's possible to physically
insert 1000BASE-LX10 SFP's besides each other? Any experience with
this?


In addition, as I checked the 4900 series as well, do built in X2
ports(the ones in the chassis http://mcaf.ee/us1dt) in 4900M support
TwinGig Converter Module? According to documentation they don't. Any
experience with this?


regards,
martin


2011/10/1 Nikolay Shopiksho...@inblock.ru:

Martin,

C3KX-NM-10GT allow you to install only two 10Gb SFP+ or 4 1Gbit SFP, or just
ony 10Gbit SFP+ plus 2 1Gbit SFP. You can take look here
http://www.cisco.com/en/US/prod/collateral/switches/ps5718/ps6406/data_sheet_c78-584733.html
at table 4.



On 30.09.2011 23:44, Martin T wrote:


Alan, Arne:

WS-C3750X-24S-E + C3KX-NM-10GT seems to be a perfect solution!

WS-C3750X-24S-E has 24 SFP ports so one could use 12 1000BASE-T SFP's
+ 2 1000BASTE-LX10 SFP's and C3KX-NM-10GT module provides two
10GBASE-T ports so all the requirements seems to be fulfilled.


Jim,
I'm afraid WS-C4928-10GE would not suite because two 10GBASE-T ports
are must-have. If there would be 10GBASE-T X2 modules available, this
switch would be a nice option.


Kevin,
if possible, I would prefer Cisco.



What do you think about WS-C3750X-24S-E + C3KX-NM-10GT solution? In
addition, are there 10GBASE-T SFP+ modules available on the
market(Cisco or third-party)?


regards,
martin


2011/9/30 Kevin Lochkl...@kl.net:


Martin T wrote:


Is there a Cisco switch(non-modular preferably) which fulfils those
requirements:

1) 12 or more 1000BASE-T ports
2) 2x SFP port for 1000BASE-LX10 SFP's
3) 2x 10GBASE-T ports(for IBM 10Gb iSCSI Host Interface Card 81Y9613,
which has two 10GBASE-T interfaces)


It's not Cisco but Dell 6248 does have a 2 port 10Gbase-T module
option.  The onboard SFP ports will accept any generic 1G SFP's
(for your LX10 optics).

- Kevin

___
cisco-nsp mailing list  cisco-nsp@puck.nether.net
https://puck.nether.net/mailman/listinfo/cisco-nsp
archive at http://puck.nether.net/pipermail/cisco-nsp/



___
cisco-nsp mailing list  cisco-nsp@puck.nether.net
https://puck.nether.net/mailman/listinfo/cisco-nsp
archive at http://puck.nether.net/pipermail/cisco-nsp/



___
cisco-nsp mailing list  cisco-nsp@puck.nether.net
https://puck.nether.net/mailman/listinfo/cisco-nsp
archive at http://puck.nether.net/pipermail/cisco-nsp/


[c-nsp] pppoe-link

2011-10-02 Thread K bharathan
| upstream ISP || Fibre Link |--| downstream ISP (gw
rt) |

Fibre link (1 line) is divided into 3 VLANS
1 vlan for internet bandwidth (wan ip from upstream ISP
1 vlan for DSL bandwidth (wan ip from upstream ISP)
1 vlan (l2tp link) for PPPOE from their VPLS network; no wan ips on this;

the link terminates at downstream ISP gw router

problem: there is no wan ip on the PPPoe link how a PPPoe server can  be
answering the DSL client; how can this particular link can be  connected to
the downstream ISP network; anybody has got any ideas ?
thanks
___
cisco-nsp mailing list  cisco-nsp@puck.nether.net
https://puck.nether.net/mailman/listinfo/cisco-nsp
archive at http://puck.nether.net/pipermail/cisco-nsp/


Re: [c-nsp] pppoe-link

2011-10-02 Thread Arie Vayner (avayner)
I am not completely sure about the PPPoE model you have as you mention 2
different things:
- L2TP
- VPLS

If you get a L2 circuit (VPLS) with no IP configured on it, and you
terminate PPPoE sessions on your router, then this is not L2TP, but
regular PPPoE.
PPPoE is a L2 protocol which runs directly over Ethernet and does not
require an IP layer (the customer does not get an IP address before the
PPP layer assigns it).

With L2TP you would have an IP based tunnel from the SP, while they
terminate the PPPoE session, and they would tunnel it to your LNS. In
that case the LAC (SP BRAS) has to have IP connectivity to your LNS
device.

Arie

-Original Message-
From: cisco-nsp-boun...@puck.nether.net
[mailto:cisco-nsp-boun...@puck.nether.net] On Behalf Of K bharathan
Sent: Sunday, October 02, 2011 11:40
To: cisco-nsp@puck.nether.net
Subject: [c-nsp] pppoe-link

| upstream ISP || Fibre Link |--| downstream ISP 
| (gw
rt) |

Fibre link (1 line) is divided into 3 VLANS
1 vlan for internet bandwidth (wan ip from upstream ISP
1 vlan for DSL bandwidth (wan ip from upstream ISP)
1 vlan (l2tp link) for PPPOE from their VPLS network; no wan ips on
this;

the link terminates at downstream ISP gw router

problem: there is no wan ip on the PPPoe link how a PPPoe server can  be
answering the DSL client; how can this particular link can be  connected
to the downstream ISP network; anybody has got any ideas ?
thanks
___
cisco-nsp mailing list  cisco-nsp@puck.nether.net
https://puck.nether.net/mailman/listinfo/cisco-nsp
archive at http://puck.nether.net/pipermail/cisco-nsp/

___
cisco-nsp mailing list  cisco-nsp@puck.nether.net
https://puck.nether.net/mailman/listinfo/cisco-nsp
archive at http://puck.nether.net/pipermail/cisco-nsp/


Re: [c-nsp] Huawei NE40E-X3 vs Cisco AS9K

2011-10-02 Thread krunal shah
Consider Brocade MLX or MLXe 
Krunal


On Wed, Sep 28, 2011 at 2:56 AM, Manuel Marín m...@transtelco.net wrote:

 Hi

 We are currently looking for alternatives to upgrade cisco 76XX  routers
 and
 we are comparing Huawei NE40E-X3 vs Cisco ASR9K. I was wondering if someone
 can share their experience with Huawey routers as Core MPLS routers.

 Any advice would be greatly appreciated

 Thanks
 ___
 cisco-nsp mailing list  cisco-nsp@puck.nether.net
 https://puck.nether.net/mailman/listinfo/cisco-nsp
 archive at http://puck.nether.net/pipermail/cisco-nsp/

___
cisco-nsp mailing list  cisco-nsp@puck.nether.net
https://puck.nether.net/mailman/listinfo/cisco-nsp
archive at http://puck.nether.net/pipermail/cisco-nsp/


[c-nsp] C7600 vs. ASR 9000

2011-10-02 Thread Michał Grzybczyk

Hi,

Any opinion which one is the better choice for a core network ?
How much different is OS on ASR 9000 in comparizon to IOS ?
If I know IOS there is no problem to operate it ?

Assuming that I need MPLS, VPLS, H-QOS ...
and of course the most important is stable, stable ... and reliable
device !


Thanks in advance,
Regards,
grzybek

___
cisco-nsp mailing list  cisco-nsp@puck.nether.net
https://puck.nether.net/mailman/listinfo/cisco-nsp
archive at http://puck.nether.net/pipermail/cisco-nsp/


Re: [c-nsp] pppoe-link

2011-10-02 Thread K bharathan
On Sun, Oct 2, 2011 at 2:48 PM, Arie Vayner (avayner) avay...@cisco.comwrote:

 I am not completely sure about the PPPoE model you have as you mention 2
 different things:
 - L2TP
 - VPLS

 If you get a L2 circuit (VPLS) with no IP configured on it, and you
 terminate PPPoE sessions on your router, then this is not L2TP, but
 regular PPPoE.
 PPPoE is a L2 protocol which runs directly over Ethernet and does not
 require an IP layer (the customer does not get an IP address before the
 PPP layer assigns it).

 With L2TP you would have an IP based tunnel from the SP, while they
 terminate the PPPoE session, and they would tunnel it to your LNS. In
 that case the LAC (SP BRAS) has to have IP connectivity to your LNS
 device.

 Arie

 -Original Message-
 From: cisco-nsp-boun...@puck.nether.net
 [mailto:cisco-nsp-boun...@puck.nether.net] On Behalf Of K bharathan
 Sent: Sunday, October 02, 2011 11:40
 To: cisco-nsp@puck.nether.net
 Subject: [c-nsp] pppoe-link

 | upstream ISP || Fibre Link |--| downstream ISP
 | (gw
 rt) |

 Fibre link (1 line) is divided into 3 VLANS
 1 vlan for internet bandwidth (wan ip from upstream ISP
 1 vlan for DSL bandwidth (wan ip from upstream ISP)
 1 vlan (l2tp link) for PPPOE from their VPLS network; no wan ips on
 this;

 the link terminates at downstream ISP gw router

 problem: there is no wan ip on the PPPoe link how a PPPoe server can  be
 answering the DSL client; how can this particular link can be  connected
 to the downstream ISP network; anybody has got any ideas ?
 thanks
 ___
 cisco-nsp mailing list  cisco-nsp@puck.nether.net
 https://puck.nether.net/mailman/listinfo/cisco-nsp
 archive at http://puck.nether.net/pipermail/cisco-nsp/


thanks

but what is the difference between regular PPPOE and L2TP; it seems i've to
get more clarity from SP; but this given VLAN doe sn't have any IP; find
difficulty to set a gw for
the PPPoe server (NAS on cisco 2900 ISR router)sitting in the LAN
___
cisco-nsp mailing list  cisco-nsp@puck.nether.net
https://puck.nether.net/mailman/listinfo/cisco-nsp
archive at http://puck.nether.net/pipermail/cisco-nsp/


[c-nsp] Large number of arp entries on 2960G

2011-10-02 Thread John Elliot

Hi Guys,
Running a management vlan(11) on a 2960S stack-2960G-7200 +  2509(for OOB) - 
i.e. 4 IP's
sh arp on 2960s, shows 3 entries (int vlan11)sh arp on 2509, shows 3 entries 
(int eth0)sh arp on 7200, shows 4 entries(on dot1q vlan 11)sh arp on 2960g, 
shows over 1000 entries, all with the 7200's mac address, all on interface 
vlan11 - all entries appear to be random IP's, in that they are routes(IP's) 
learned from upstream bgp peering sessions and also some from our internal 
ospf...none of these bgp sessions or ospf are running in dot1q vlan11
The only difference I can see on the 2 switches vlan interfaces is the 
2960g(The one with all the strange arp entries) has no ip route-cache.
Any suggestions as to what could be causing this?
Cheers.



  
___
cisco-nsp mailing list  cisco-nsp@puck.nether.net
https://puck.nether.net/mailman/listinfo/cisco-nsp
archive at http://puck.nether.net/pipermail/cisco-nsp/


Re: [c-nsp] Large number of arp entries on 2960G (Solved)

2011-10-02 Thread John Elliot

Found the issue - The 2960g was missing default gw of the 7200.

 From: johnellio...@hotmail.com
 To: cisco-nsp@puck.nether.net
 Date: Mon, 3 Oct 2011 13:09:49 +1100
 Subject: [c-nsp] Large number of arp entries on 2960G
 
 
 Hi Guys,
 Running a management vlan(11) on a 2960S stack-2960G-7200 +  2509(for OOB) 
 - i.e. 4 IP's
 sh arp on 2960s, shows 3 entries (int vlan11)sh arp on 2509, shows 3 entries 
 (int eth0)sh arp on 7200, shows 4 entries(on dot1q vlan 11)sh arp on 2960g, 
 shows over 1000 entries, all with the 7200's mac address, all on interface 
 vlan11 - all entries appear to be random IP's, in that they are 
 routes(IP's) learned from upstream bgp peering sessions and also some from 
 our internal ospf...none of these bgp sessions or ospf are running in dot1q 
 vlan11
 The only difference I can see on the 2 switches vlan interfaces is the 
 2960g(The one with all the strange arp entries) has no ip route-cache.
 Any suggestions as to what could be causing this?
 Cheers.
 
 
 
 
 ___
 cisco-nsp mailing list  cisco-nsp@puck.nether.net
 https://puck.nether.net/mailman/listinfo/cisco-nsp
 archive at http://puck.nether.net/pipermail/cisco-nsp/
  
___
cisco-nsp mailing list  cisco-nsp@puck.nether.net
https://puck.nether.net/mailman/listinfo/cisco-nsp
archive at http://puck.nether.net/pipermail/cisco-nsp/


Re: [c-nsp] Large number of arp entries on 2960G

2011-10-02 Thread Ian Henderson
On 03/10/2011, at 1:09 PM, John Elliot wrote:

 interface vlan11 - all entries appear to be random IP's, in that they are 
 routes(IP's) learned from upstream bgp peering sessions and also some from 
 our internal ospf...none of these bgp sessions or ospf are running in dot1q 
 vlan11

Smells like one of the devices is doing Proxy ARP. This is usually bad, 
particularly if its trying to ARP for all hosts on the Internet - will drive up 
CPU and memory usage.


___
cisco-nsp mailing list  cisco-nsp@puck.nether.net
https://puck.nether.net/mailman/listinfo/cisco-nsp
archive at http://puck.nether.net/pipermail/cisco-nsp/


Re: [c-nsp] Large number of arp entries on 2960G (Solved)

2011-10-02 Thread John Elliot


Thanks(everyone) - The 7200 has a number of dot1q subints, can I disable proxy 
arp on the physical gig int(i.e. will this apply the setting to all the dot1q 
ints), or must it be disabled on all dot1q subints?



 Subject: Re: [c-nsp] Large number of arp entries on 2960G (Solved)
 From: l...@cisco.com
 Date: Mon, 3 Oct 2011 13:27:08 +1100
 CC: cisco-nsp@puck.nether.net
 To: johnellio...@hotmail.com
 
 On 03/10/2011, at 1:23 PM, John Elliot wrote:
 
  Found the issue - The 2960g was missing default gw of the 7200.
 
 suggest you disable proxy arp on whatever the offending device(s) are too.
 
 thats papering over what the root problem is..
 
 
 cheers,
 
 lincoln.
 
  
___
cisco-nsp mailing list  cisco-nsp@puck.nether.net
https://puck.nether.net/mailman/listinfo/cisco-nsp
archive at http://puck.nether.net/pipermail/cisco-nsp/


Re: [c-nsp] Cisco and third party transceivers

2011-10-02 Thread Dennis
On Tue, Sep 27, 2011 at 1:17 PM, Martin T m4rtn...@gmail.com wrote:
 Hello,
 there are providers like Flexoptix(http://www.flexoptix.net) who are
 able to flash SFP EEPROM memory with different vendor data, probably
 set custom serial numbers etc. However, why is such service needed at
 nowadays for Cisco equipment? Every router/switch I have seen supports
 the service unsupported-transceiver(should turn of checking the
 Cisco Quality ID) and no errdisable detect cause gbic-invalid
 commands and thus doesn't check the ID code in EEPROM..
 In addition, am I correct, that some old Cisco IOS versions did not
 have the service unsupported-transceiver and no errdisable detect
 cause gbic-invalid commands and thus one really was forced to use
 transceivers with Cisco serials?


Some ASAs won't take some 3rd party transceivers no matter what voodoo
you type, as far as I've been able to figure out anyway.  It really
depends on your business model so to speak.
___
cisco-nsp mailing list  cisco-nsp@puck.nether.net
https://puck.nether.net/mailman/listinfo/cisco-nsp
archive at http://puck.nether.net/pipermail/cisco-nsp/