Skip to content

ANFTechTeam/accelnet_tuning

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

6 Commits
 
 
 
 

Repository files navigation

accelnet_tuning

Network interface cards deliver the best performance when associated with the lowest number set of physical cores on a compute instance. As physical cores are bundled by NUMA node, it is best to associate the network interface with the lowest numbered NUMA node. Hypervisors cannot always be counted on to perform NUMA node mapping in physical order, this tool was developed with the cooperation of both the Azure Linux User Experience team and the Azure NetApp Files team to identify and set the accelerated networking interface on the best possible NUMA node for a given compute instance.

Note: This tool should be run each time that a machine is started as NUMA mapping is only consistant while a machine is powered up.

What it does:

  1. The tool will take as input a file name placed on an NFS mounted filesystem:

    1a) ./set_best_affinity.sh /nfs-mount-point/testfile

  2. The file /nfs-mount-point/testfile is created 8GiB in size if it does not already exist

  3. The tool will on its own identify the accelerated network interface through which the NFS mount is accessed, let us say for example eth3

  4. The tool will set the number of tx/rx queues associated with the to which ever is less, the number of cores on a single NUMA node or nic maximum.

  5. The tool will associate eth3 one by one with each of the NUMA nodes on the system and run 5 sets of dd read commands tabulating the throughput of each run as well as the series average per NUMA node.

  6. After all tests are done, the tool will associate the accelerated networking interface with the most optimal NUMA node.

About

No description, website, or topics provided.

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages