Re: [chrony-dev] [PATCH] MacOS X - Increase priority of chronyd process |
[ Thread Index |
Date Index
| More chrony.tuxfamily.org/chrony-dev Archives
]
- To: chrony-dev@xxxxxxxxxxxxxxxxxxxx
- Subject: Re: [chrony-dev] [PATCH] MacOS X - Increase priority of chronyd process
- From: Bryan Christianson <bryan@xxxxxxxxxxxxx>
- Date: Mon, 24 Aug 2015 19:53:11 +1200
- Dkim-signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=smtpcorp.com; s=a0-2; h=Feedback-ID:X-Smtpcorp-Track:To:Message-Id:Date: From:Subject; bh=/dliBXTPsmdkhy0Nsl4Ao5MH0+vJO1bzRP5sRioQPUo=; b=FVQkDeCf463c GiL7MKtQEhHEJKKQkCLtezrHVeRk2CBxBiXVpm5UG0UtCFZ4dMJOEdBmiFhbZNO7WTuu4MeU3Vov6 vY+FXycEVdwsQow9WUWqdbuOt3SuHvM6M787RwsN6aQmf+L7lvUtCAxm09XFvpsU/mu4R1gXH2rmo djGg/gmSp89UyKyLdrCb//XqJqM09ozVFCQ5BOCt9Vvjt+FwD9dpoXp9XnAN2B15n9O4NipiVnNdo W8MaCBweqiMygN9RjTc+nEbvepFRXy8KEsZzQRaFYMDYCFbjnHAmfuKKmiUbQ8CTbvQUL43DLo5FG pMvxSPn+W5XF4+fLq5gW3g==;
- Feedback-id: 149811m:149811acx33YQ:149811szkwWYF0rr:SMTPCORP
> On 24/08/2015, at 11:06 am, Bill Unruh <unruh@xxxxxxxxxxxxxx> wrote:
>
> Are you sure that there is a 1us offset? How do you measure that? Remember
> that anything in the computer has latency and 1us is not that long a time. Ie, it takes a bit of work to measure that offset to that accuracy.
I am reading the 'Rem. corr.' value from /var/log/chrony/tracking.log
running this through a low pass filter and plotting the result. The graphs clearly show approximately 1 to 2 us offset when running chronyd at default priority. There are couple of graphs at my website that show the effect of enabling real-time scheduling.
I'm not doing any measurement per se, just relying on the output stats from chronyd.
The low pass filter in my graphing software is a simple exponential weighting as used elsewhere in chronyd:
mean_corr = 0.1 * (corr - mean_corr)
>
> Note that the only time critical item is the reading of the system clock when
> the interrupt comes in. And I have doubts that the interrupt handling software
> is up to much less than 1us standard.
You may be right on that - I don't know the granularity of interrupt timings in the MacOS kernel running on an i7 processor.
One other thing - I have not seen any deterioration in the performance of other processes caused by chronyd running as a real-time process and chronyd still uses minimal cpu and memory.
--
Bryan Christianson
--
To unsubscribe email chrony-dev-request@xxxxxxxxxxxxxxxxxxxx with "unsubscribe" in the subject.
For help email chrony-dev-request@xxxxxxxxxxxxxxxxxxxx with "help" in the subject.
Trouble? Email listmaster@xxxxxxxxxxxxxxxxxxxx.