Distributed Systems 3.1: Physical time
Vložit
- čas přidán 27. 07. 2024
- Accompanying lecture notes: www.cl.cam.ac.uk/teaching/212...
Full lecture series: • Distributed Systems le...
This video is part of an 8-lecture series on distributed systems, given as part of the undergraduate computer science course at the University of Cambridge. It is preceded by an 8-lecture course on concurrent systems for which videos are not publicly available, but slides can be found on the course web page: www.cl.cam.ac.uk/teaching/212...
lol the part at 17:37
Thank you for providing these lectures. As just a random software engineer who is currently trying to learn about distributed system, it really helps me.
This is an awesome series of lectures. I really wish I had you as professor during my university time!
I knew nothing abt this until today 😂😂😂 What a lecture 👏🏻👏🏻
I guess you can say that smearing is a hack, but I would say that leap seconds are a hack. Maybe they should have done smearing in the first place. Of course, both have disadvantages, but I think most applications would prefer that each day has the same number of seconds as opposed to wanting that each second has the same length.
Awesome video!
It's really interesting how the theoretical definition of time deviates from the that derived from the physical world.
great lecture
My first though was that the 30 june 2012 outage happened due to the unix timestamp value exceeding the limits of int32, but I checked and that won't happen till 2038. Probably something to do with the fact that 2012 was a leap year - I know that throws off clocks. One of the best videos on the issues of dealing with clocks in code is this Computerphile video from so long ago: czcams.com/video/-5wpm-gesOY/video.html
Great lecture. Just curious about the synchronization of atomic clocks in GPS satellites & earth. Do we have any effect due to relativity? Doesn't relative motion stretch the time a little bit?
Good question! I was about to say that it will be negligible, but if you think about it, you do see things like stars changing their positions during a total solar eclipse, so you don't necessarily need to be near a black hole to start noticing things.
Hello Martin thanks for the series
Not sure what you are looking for, but there are links in the video's description.
The materials are here: www.cl.cam.ac.uk/teaching/2122/ConcDisSys/materials.html but the videos are only available to Cambridge students. The concurrent systems course is given by a different lecturer, and it's each lecturer's own decision whether to make the recordings publicly available.
@@kleppmann Thank you Martin!
i want to see a computer time zones collaboration with tom scott
this aged well
If we are making adjustments to atomic clock time based on earths natural events, then why do we not directly use time based on astronomy? Can anyone explain?
An atomic clock fits in a small box, works anytime, and provides time with very high resolution (nanoseconds). Astronomical observations require a telescope pointed at the stars, don't work when it's cloudy, and are slow and expensive.
So then basically, eventually the bug that came in the past due to the extra second can repeat again eventually right? Or was there any lesson learnt and is guaranteed to not repeat?
I tried finding an answer to this online but couldn't find anything.