TLK10031 HS_ERROR_CNT

we use the Disable 10GBASE-R FEC function 

Hi 

we use iperf for udp packet test

But i find the HS_ERROR_COUNT description is as followed

it is also valid in function mode 

Hi,

Thanks for the correction, you are right that HS_ERROR_COUNT reports invalid code words received by the decoder.  It seems odd that you're observing errors without observing any packet loss.  I will check with a colleague regarding this and get back to you.

Thanks,
Drew

Hi,

Apologies for the delay on this, I was out of office over the last week.

I have a couple follow up questions related to this.

• I'm assuming data is 10G Ethernet 64B/66B encoded?
• Is there any consistency in the number of errors you are observing?  Are the errors static (e.g. a few errors during init), or do the errors accumulate over time?
• Are there any test data formatting/settings you might be able to share, such as packet size and utilization rate?  Is there any correlation between data format and errors?

Thanks,

Drew

HI 

. yes , i am sure that it is 64b/66b encoded

. the observing time is at any time we tesing. for example  when link up is ok, we read the cnt , and then for some times later we read again.

. we use the iperf tool and the udp packet size is 1470 bytes. we use 10g-base r mode (#disable auto-negotiation；#turn off lt_training) the utilization  rate is full rate 10g mode.  

Hi,

How does the error count vary over time?  Does the error count continue to increase, or does it remain static?

Thanks,
Drew