Puzzling clock offset spikes

[Paul Theodoropoulos]

On 6/30/2018 10:00 AM, Paul Theodoropoulos wrote:
> Yeah, I definitely need to try this. I know that any significant fs 
> activity will cause spikes. And there's a nice setup out there that 
> will write to memory fs, and flush back to disk when rebooting. I've 
> tried adjusting the 'commit' interval both longer (60s) and shorter 
> (5s) but no apparent difference.
>
  A little more info -

I've futzed around with vm.vfs_cache_pressure, 
vm.dirty_background_ratio, vm.dirty_ratio, vm.dirty_writeback_centisecs, 
all to no apparent difference.

I'm on an asymmetric speed WAN connection, 55Mbps down, 6Mbps up, 
however my latency is not bad, about 24ms ping RTT to most of my peers. 
I don't see that as a factor, and as mentioned, I've sniffed packets and 
seen no untoward spikes of traffic.

What else...boot sdhc is class 10, all good specs otherwise. I have the 
mmc0 bus overclocked from 50MHz to 100MHz, but the spikes long predate 
that change.


-- 
Paul Theodoropoulos
www.anastrophe.com