Re: [Xen-devel] [PATCH 1/3] Make credit2 the default scheduler

[George Dunlap <george.dunlap@xxxxxxxxxx>]

On 02/05/2018 11:45 AM, Andrew Cooper wrote:
> On 05/02/18 11:36, George Dunlap wrote:
>> Credit2 was declared "supported" in 4.8, and as of 4.10 had two other
>> critical features implemented (soft affinity / NUMA and caps).
>>
>> Signed-off-by: George Dunlap <george.dunlap@xxxxxxxxxx>
> 
> On what justification?
> 
> Switching the default involves demonstrating that the new default is
> better than the old one, and simply "we've implemented more features"
> isn't good enough IMO.

Well the code is certainly better: more predictable, less jitter, easier
to determine how modifications will affect overall behavior, easier in
the future to make load-balancing behavior more subtle (e.g., taking
into account the cost of powering up extra cores, &c).

> The last set of measurements I recall seeing said that credit1 and
> credit2 were mostly kneck-and-kneck, but there was still a noticeable
> difference in credit1's favour when it came to aggregate small-packet
> TCP throughput, and no benchmark where credit2 was clearly a winner.

It's somewhat of a mixed bag.  Unfortunately most of what I have are
tests using XenServer's internal perf testing system, so I can't share
the raw data (via links anyway).

Here is a summary of data from an intern e-mail Dario sent about 6
months ago:

* DVDbench: On underloaded systems, credit2 outperformed credit1 by
about 4%.  On overloaded systems, credit2 underperformed by about 3%.

* On a range of tests (unixbench, lmbench, &c), credit and credit2
perform within 5% of each other (up and down).

* Credit2 fairly consistently beats credit for TCP-style workloads.

* Credit2 is sometimes equal to, sometimes 5-15% worse than, credit for
synthetic CPU workloads (e.g., Dhrystone).

* On LoginVSI, credit2 fairly consistently outperforms credit by about 10%.

Credit2, like credit, has a number of workloads / setups for which
performance could be improved.  Personally I think networking and
partially-loaded systems is going to be more representative of what Xen
is actually used for; so I think credit2 is on the whole the better
scheduler to use by default.  And in any case, making those improvements
on credit2 will be easier than on credit.

 -George

_______________________________________________
Xen-devel mailing list
Xen-devel@xxxxxxxxxxxxxxxxxxxx
https://lists.xenproject.org/mailman/listinfo/xen-devel