[Ndn-lib] PSync in CCL
jefft0 at remap.ucla.edu
Mon Jan 28 13:50:02 PST 2019
For us, the driver for implementing PSync in the CCL is to use it in the Common Name Library to synchronize sets of names. I’m almost finished with Full PSync in NDN-CPP. (This week.) Next would be in PyNDN to use in the Python version of the Common Name Library. (I am not sure of the available implementations of MurmurHash in Python. If it’s easy, then it’s only a few days, otherwise I have to implement and test.)
Once we have Full PSync in NDN-CPP and PyNDN, it should be very easy to add Partial PSync. I’ll be interested to see the discussion for your question about the difference between Partial and Full PSync, and your use cases. For the Common Name Library, we want Full because it uses multicast and is better for when new producers are able to join the sync group. (For some reason that Ashlesh can explain, Full also uses a different Bloom Filter implementation.)
- Jeff T
From: Junxiao Shi <shijunxiao at email.arizona.edu<mailto:shijunxiao at email.arizona.edu>>
Date: Monday, January 28, 2019 at 13:07:00
To: NDN Lib <ndn-lib at lists.cs.ucla.edu<mailto:ndn-lib at lists.cs.ucla.edu>>
Cc: Jeff Thompson <jefft0 at remap.ucla.edu<mailto:jefft0 at remap.ucla.edu>>
Subject: PSync in CCL
I heard in NFD call that you plan to adopt PSync protocol in Common Client Libraries. I wonder what's the timeline for this?
Please break down the timeline answers to (1) date for API design (2) date for each library.
-------------- next part --------------
An HTML attachment was scrubbed...
More information about the Ndn-lib