DevinLeamy opened a new pull request, #483:
URL: https://github.com/apache/mesos/pull/483

   Currently egress rate limit scaling requires manual configuration of per CPU 
rate limit. While it is more predictable and gives rounder values, in a 
heterogeneous environment where hosts with one NIC type have different CPUs it 
may be cumbersome to configure.
   
   This commit allows you to pass an "auto" value to the --egress_rate_per_cpu 
flag of the isolator
   ```
   --egress_rate_per_cpu=auto
   ```
   to automatically calculate the value. The value is calculated by determining 
the link speed using the /sys/class/net/<iface>/speed interface 
(https://www.kernel.org/doc/Documentation/ABI/testing/sysfs-class-net) and 
dividing the speed by the number of CPUs resources that are available.  
--egress_rate_per_cpu=auto will fail if there are 0 CPU resources.
   
   The `network_link_speed` flag
   ```
   --network_link_speed=<link speed in bytes/second>
   ```
   can be used to override link speed detection when --egress_rate_per_cpu=auto.


-- 
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.

To unsubscribe, e-mail: [email protected]

For queries about this service, please contact Infrastructure at:
[email protected]

Reply via email to