cancel
Showing results forĀ 
Search instead forĀ 
Did you mean:Ā 
cancel
1013
Views
0
Helpful
3
Replies

Regarding Ping Latency in 5548UP Model with Daughter card L3.

Rishav Garg
Level 1
Level 1

Hi All,

I have a 5548UP model in data center as L3 working. Now my question is when i generate a ping from Switch itself to down host. I am getting latency around 2 to 8 Ms and if i generate a ping from down host to other host in same or another subnet latency would be <1ms.

And in cisco documentation i read port to port latency is 1.8 ms and with layer 3 Daughter card i.e approx 3.8 ms. Can somebody explain why i am getting <1ms latency from down host. No matter another host is in same or another subnet. And why ( around 2 to 8 Ms) from switch itself. 

If it so then what is the exact meaning that in documentation as above. 

 

Regards

Rishav

Data Center Engineer

1 Accepted Solution

Accepted Solutions

Jon Marshall
Hall of Fame
Hall of Fame

Rishav

The usual reason with switches is the pings through the switch ie. end device to another end device are simply hardware switched, no need to involve the main CPU.

A ping to the switch itself is different though because that has to go up the protocol stack on the switch and usually this is done by the main CPU.

So whenever you test latency it is always best with switches to ping through rather than to the switch.

There are times when the main CPU of the switch can be busy and your ping times are quite high but this does not reflect how well it actually switching packets through from one device to another.

The above is a general answer in terms of switches although I expect it applies to Nexus switches as well.

Jon

View solution in original post

3 Replies 3

Jon Marshall
Hall of Fame
Hall of Fame

Rishav

The usual reason with switches is the pings through the switch ie. end device to another end device are simply hardware switched, no need to involve the main CPU.

A ping to the switch itself is different though because that has to go up the protocol stack on the switch and usually this is done by the main CPU.

So whenever you test latency it is always best with switches to ping through rather than to the switch.

There are times when the main CPU of the switch can be busy and your ping times are quite high but this does not reflect how well it actually switching packets through from one device to another.

The above is a general answer in terms of switches although I expect it applies to Nexus switches as well.

Jon

I totally agree with you.

But i want to know what's the meaning whatever document says like " port to port latency is 1.8 ms without going upto Layer3 means Purely L2 switching and with layer 3 Daughter card i.e approx 3.8 ms". 

Is it means max would be 1.8ms or something else. Kindly explain.

Refrence Words : 

The Layer 3 card can forward up to 160 Gbps of traffic, or 240 million packets per second (mpps).

The First In, First Out (FIFO) port-to-port latency for the Cisco Nexus 5548P is approximately 2 microseconds, and for the Cisco Nexus 5548UP and 5596UP it is approximately 1.8 microseconds. With the Layer 3 card, the FIFO port-to-port latency for routed traffic is approximately 4.7 microseconds.
Link: http://www.cisco.com/c/en/us/products/collateral/switches/nexus-5000-series-switches/guide_c07-673997.html

Hi Jon

 

Kindly reply to my latest post.

I will be very thankful to you.

 

Rishav

Review Cisco Networking for a $25 gift card