cancel
Showing results for 
Search instead for 
Did you mean: 
cancel
1146
Views
0
Helpful
7
Replies

I need to upgrade my WS-X6148-GE-TX cards to something with more throughput for server farms.

markmarek
Level 1
Level 1

The WS-X6148-GE-TX is more of a desktop solution. These cards are dropping packets and I need to have them replaced to cards that support server farms. What should I upgrade to?

2 Accepted Solutions

Accepted Solutions

Giuseppe Larosa
Hall of Fame
Hall of Fame

Hello Mark,

to get an appropriate answer you need to provide a sh module of your C6500. Actually as explained in a very recent thread not all combinations of linecards and supervisor models are possible.

a good starting point for you is the C6500 architecture whitepaper

http://www.cisco.com/en/US/prod/collateral/switches/ps5718/ps708/prod_white_paper0900aecd80673385.html

so for example if you have a sup720 and enough budget you can buy WS-6748.

But if you haven't a sup720 you cannot use them

Cisco Catalyst 6500 Architecture: CEF720 Line Cards

In 2003, Cisco announced the next generation Supervisor Engine 720 along with a new series of line cards. They are designed to take advantage of architecture extensions of the Supervisor Engine 720 and the enhanced crossbar switch fabric.
Four Cisco Express Forwarding 720 line cards are available providing higher performance, and higher port densities of high speed interfaces over earlier generations of line cards. The new CEF720 line cards (Figure 18) are:

WS-X6724-SFP: a 24 port Gigabit Ethernet SFP based line card supporting a single 20-Gbps fabric channel to the Supervisor Engine 720 crossbar switch fabric. Also supports an optional Distributed Forwarding Card 3a (DFC3a, DFC3B or DFC3BXL)

WS-X6748-GE-TX: a 48 port 10/100/1000 RJ45 based line card supporting 2 x 20-Gbps fabric channels to the Supervisor Engine 720 crossbar switch fabric. Also supports an optional Distributed Forwarding Card 3a (DFC3a, DFC3B or DFC3BXL)

WS-X6748-SFP: a 48 port 1000Mb SFP based line card supporting 2 x 20-Gbps fabric channels to the Supervisor Engine 720 crossbar switch fabric. Also supports an optional Distributed Forwarding Card 3 (DFC3a, DFC3B or DFC3BXL)

Hope to help

Giuseppe

View solution in original post

markmarek wrote:

I do not have vpn access to the switch at the moment. But I do have WS-SUP720-3B's.

Mark

Then as Giuseppe said you can use the 67xx modules. If you are looking for an RJ45 10/100/1000 ethernet module for servers you should go for the WS-X6748-GE-TX. If you need a module that terminates fibre then the WS-X6724-SFP/WS-X6748-SFP.

The 6748 modules have a 2 x 20Gbps connections to the switch fabric so it is hard to actually oversubscribe the card ie. you would need more than 40 servers all putting out 1Gbps to oversubscribe it.

The 6724 module has 1 x 20Gbps to the switch fabric.

All of the above modules are made for either interconnecting switches or for server farms.

If budget allows these are the modules you should be considering if throughput is your major concern.

Jon

View solution in original post

7 Replies 7

Giuseppe Larosa
Hall of Fame
Hall of Fame

Hello Mark,

to get an appropriate answer you need to provide a sh module of your C6500. Actually as explained in a very recent thread not all combinations of linecards and supervisor models are possible.

a good starting point for you is the C6500 architecture whitepaper

http://www.cisco.com/en/US/prod/collateral/switches/ps5718/ps708/prod_white_paper0900aecd80673385.html

so for example if you have a sup720 and enough budget you can buy WS-6748.

But if you haven't a sup720 you cannot use them

Cisco Catalyst 6500 Architecture: CEF720 Line Cards

In 2003, Cisco announced the next generation Supervisor Engine 720 along with a new series of line cards. They are designed to take advantage of architecture extensions of the Supervisor Engine 720 and the enhanced crossbar switch fabric.
Four Cisco Express Forwarding 720 line cards are available providing higher performance, and higher port densities of high speed interfaces over earlier generations of line cards. The new CEF720 line cards (Figure 18) are:

WS-X6724-SFP: a 24 port Gigabit Ethernet SFP based line card supporting a single 20-Gbps fabric channel to the Supervisor Engine 720 crossbar switch fabric. Also supports an optional Distributed Forwarding Card 3a (DFC3a, DFC3B or DFC3BXL)

WS-X6748-GE-TX: a 48 port 10/100/1000 RJ45 based line card supporting 2 x 20-Gbps fabric channels to the Supervisor Engine 720 crossbar switch fabric. Also supports an optional Distributed Forwarding Card 3a (DFC3a, DFC3B or DFC3BXL)

WS-X6748-SFP: a 48 port 1000Mb SFP based line card supporting 2 x 20-Gbps fabric channels to the Supervisor Engine 720 crossbar switch fabric. Also supports an optional Distributed Forwarding Card 3 (DFC3a, DFC3B or DFC3BXL)

Hope to help

Giuseppe

markmarek
Level 1
Level 1

I do not have vpn access to the switch at the moment. But I do have WS-SUP720-3B's.

markmarek wrote:

I do not have vpn access to the switch at the moment. But I do have WS-SUP720-3B's.

Mark

Then as Giuseppe said you can use the 67xx modules. If you are looking for an RJ45 10/100/1000 ethernet module for servers you should go for the WS-X6748-GE-TX. If you need a module that terminates fibre then the WS-X6724-SFP/WS-X6748-SFP.

The 6748 modules have a 2 x 20Gbps connections to the switch fabric so it is hard to actually oversubscribe the card ie. you would need more than 40 servers all putting out 1Gbps to oversubscribe it.

The 6724 module has 1 x 20Gbps to the switch fabric.

All of the above modules are made for either interconnecting switches or for server farms.

If budget allows these are the modules you should be considering if throughput is your major concern.

Jon

While I wait for budget approval, which will be a few weeks, what can I do to help alleviate the symptoms for now?

Each eight ports on a 6148 share a single ASIC and 1gig's worth of bandwidth.

So take your servers that are pumping out the most traffic and spread them out

across separate 8-ports groups.  Server A in ports 1-7, Server B in ports 8-15,

Server C in ports 16-23, etc.  That will spread your heaviest used servers across

the six ASICs.  HTH

I noticed the first packet always times out the the rest are fine.

C:\Documents and Settings\Admin>ping 1.1.1.1

Pinging 1.1.1.1 with 32 bytes of data:

Request timed out.
Reply from 1.1.1.1: bytes=32 time<1ms TTL=255
Reply from 1.1.1.1: bytes=32 time<1ms TTL=255
Reply from 1.1.1.1: bytes=32 time<1ms TTL=255

Ping statistics for 1.1.1.1:
    Packets: Sent = 4, Received = 3, Lost = 1 (25% loss),
Approximate round trip times in milli-seconds:
    Minimum = 0ms, Maximum = 0ms, Average = 0ms

C:\Documents and Settings\Admin>ping 1.1.1.1

Pinging 1.1.1.1 with 32 bytes of data:

Request timed out.
Reply from 1.1.1.1: bytes=32 time<1ms TTL=255
Reply from 1.1.1.1: bytes=32 time<1ms TTL=255
Reply from 1.1.1.1: bytes=32 time<1ms TTL=255

Ping statistics for 1.1.1.1:
    Packets: Sent = 4, Received = 3, Lost = 1 (25% loss),
Approximate round trip times in milli-seconds:
    Minimum = 0ms, Maximum = 0ms, Average = 0ms

C:\Documents and Settings\Admin>ping 1.1.1.1

Pinging 1.1.1.1 with 32 bytes of data:

Request timed out.
Reply from 1.1.1.1: bytes=32 time<1ms TTL=255
Reply from 1.1.1.1: bytes=32 time<1ms TTL=255
Reply from 1.1.1.1: bytes=32 time<1ms TTL=255

Ping statistics for 1.1.1.1:
    Packets: Sent = 4, Received = 3, Lost = 1 (25% loss),
Approximate round trip times in milli-seconds:
    Minimum = 0ms, Maximum = 0ms, Average = 0ms

C:\Documents and Settings\Admin>ping 1.1.1.1

Pinging 1.1.1.1 with 32 bytes of data:

Request timed out.
Reply from 1.1.1.1: bytes=32 time<1ms TTL=255
Reply from 1.1.1.1: bytes=32 time<1ms TTL=255
Reply from 1.1.1.1: bytes=32 time<1ms TTL=255

Ping statistics for 1.1.1.1:
    Packets: Sent = 4, Received = 3, Lost = 1 (25% loss),
Approximate round trip times in milli-seconds:
    Minimum = 0ms, Maximum = 0ms, Average = 0ms

C:\Documents and Settings\Admin>ping 1.1.1.1

Pinging 1.1.1.1 with 32 bytes of data:

Request timed out.
Reply from 1.1.1.1: bytes=32 time<1ms TTL=255
Reply from 1.1.1.1: bytes=32 time<1ms TTL=255
Reply from 1.1.1.1: bytes=32 time<1ms TTL=255

Ping statistics for 1.1.1.1:
    Packets: Sent = 4, Received = 3, Lost = 1 (25% loss),
Approximate round trip times in milli-seconds:
    Minimum = 0ms, Maximum = 0ms, Average = 0ms

C:\Documents and Settings\Admin>ping 1.1.1.1

Pinging 1.1.1.1 with 32 bytes of data:

Request timed out.
Reply from 1.1.1.1: bytes=32 time<1ms TTL=255
Reply from 1.1.1.1: bytes=32 time<1ms TTL=255
Reply from 1.1.1.1: bytes=32 time<1ms TTL=255

Ping statistics for 1.1.1.1:
    Packets: Sent = 4, Received = 3, Lost = 1 (25% loss),
Approximate round trip times in milli-seconds:
    Minimum = 0ms, Maximum = 0ms, Average = 0ms

Why is the cat6509 not arp cache not working?

After the first ping the address should be cached?

show ip arp looks ok on the switches.

We did the samething for some of the servers and it helped a little but there still are unexplained latency issues. The load on the interfaces are all 1/255 but I still see lots of output drops. pings and traces show no latency but users still have to wait for email or other apps. sometimes we just flat out loses connectivity to a server and there is nothing in the switch or server logs. I am having a hard time tracking down what exactly is happening. Any advice would be appreciated. 

Getting Started

Find answers to your questions by entering keywords or phrases in the Search bar above. New here? Use these resources to familiarize yourself with the community: