LMG should test their LAN throughput differently

[Lasso1383]

Please guys don't use only SMB to test your LAN throughput. SMB transfers are influenced by a number of application-layer variables, including disk I/O performance, file system overhead, protocol inefficiencies, and even antivirus interference. You should still test SMB file transfer as it is your intended workload, but please when testing LAN throughput use something like iperf which is actually gonna work on a transport layer. Please guys I beg you please  

[Echothedolpin]

2 hours ago, Lasso1383 said:

Please guys don't use only SMB to test your LAN throughput. SMB transfers are influenced by a number of application-layer variables, including disk I/O performance, file system overhead, protocol inefficiencies, and even antivirus interference. You should still test SMB file transfer as it is your intended workload, but please when testing LAN throughput use something like iperf which is actually gonna work on a transport layer. Please guys I beg you please  

Not trying to play devil’s advocate but SMB is going to be a more realistic test for the average person than iperf. iperf will almost always set unrealistic expectations for transfer speeds whereas if you’re using a standardized image your SMB tests would show real world performance limited only by hardware (and some driver) changes.

 

Sure it’s not objective, but when the audience has the same subjective perspective it’ll be better to choose the subjective test.