Hi Barbara. Thank you for the informative video. I'm confused about your network latency calculation circa time 4:31. Doesn't the difference between the time the slave sent the delay request (108) and delay response value from the master (112) indicate the one-way network latency from slave to master? This isn't the round trip delay, so I don't know why there is a division by 2.
Hi Amit, There was initial one-way latency in the network for the original Follow Up packet, which is what we're trying to determine. The Delay Request packet is also experiencing one-way latency in the network (in the other direction). So that's why we divide by 2 - to account for the latency in both directions.
Great, is there also a video about Software PTP so with with no PTP Hardware and if using Software PTP if a acceptable choice when going for low latency in a small lab enviroment?
Cool! This protocol assumes equal links send and return as you mentioned... But is it even theoretically possible to get equal times when there is *not* equal delay for send and return?
The major use for PTP is time sensitivity below microseconds. You'll generally see applications at the service provider level (as Barbara stated), financial sector, or in the space industry. I work with a service provider and our main use for PTP is managing DOCSIS timing, since cable modems effectively "take turns" sending customer-side data to the CMTS, allowing them to occupy the same frequency without interfering with each other.
This video was very clear and thorough, it helped a lot. Thank you!
You're very welcome!
Great Work Barbara. Thank you
Thank you! I appreciate the feedback!
Hi Barbara. Thank you for the informative video. I'm confused about your network latency calculation circa time 4:31. Doesn't the difference between the time the slave sent the delay request (108) and delay response value from the master (112) indicate the one-way network latency from slave to master? This isn't the round trip delay, so I don't know why there is a division by 2.
Hi Amit, There was initial one-way latency in the network for the original Follow Up packet, which is what we're trying to determine. The Delay Request packet is also experiencing one-way latency in the network (in the other direction). So that's why we divide by 2 - to account for the latency in both directions.
Really fantastic explanation, thank you!
Thanks! I really appreciate the feedback!
Great, is there also a video about Software PTP so with with no PTP Hardware and if using Software PTP if a acceptable choice when going for low latency in a small lab enviroment?
I’m not quite sure what you’re asking but there isn’t another video. Sorry
Thank you.
You're welcome!
Best video I've ever seen! But I'm confused that this is more like a 1588 v1.
Thank you for your comment. I will look into that.
Cool! This protocol assumes equal links send and return as you mentioned...
But is it even theoretically possible to get equal times when there is *not* equal delay for send and return?
I guess it depends on the precision of your clock and the number of hops between 2 systems involved in the conversation.
Great video! thanks for sharing
Thanks for watching!
Thank YOU
Thank you too! I appreciate the feedback!
It seems like this protocol is only useful in highly dedicated networks. In other words, it wouldn’t work over the Internet.
It's really implemented in the Service Providers' networks for their transport, not as an application.
The major use for PTP is time sensitivity below microseconds. You'll generally see applications at the service provider level (as Barbara stated), financial sector, or in the space industry. I work with a service provider and our main use for PTP is managing DOCSIS timing, since cable modems effectively "take turns" sending customer-side data to the CMTS, allowing them to occupy the same frequency without interfering with each other.
Harris Robert Thomas Scott Rodriguez James