Skip to content
This repository has been archived by the owner on Apr 16, 2020. It is now read-only.

Moar benchmarks! \o/ #3

Open
daviddias opened this issue Nov 27, 2017 · 1 comment
Open

Moar benchmarks! \o/ #3

daviddias opened this issue Nov 27, 2017 · 1 comment
Labels

Comments

@daviddias
Copy link
Contributor

daviddias commented Nov 27, 2017

@goga-m, excellent work on getting this project started! I'm excited to see all the benchmarks that will follow and have the ability to visualize them!

Let's share this work on the IPFS All Hands -- ipfs/team-mgmt#528 -- today and invite everyone to contribute with proposals and other benchmarks. Off top of my head, I'm thinking:

  • file size should be adjustable so that users can experiment with different values. It is good to continue suggesting sizes (1Mb, 10Mb, 100Mb, 1Gb)
  • It should be visual from how many peers we fetched files and how much did each peer contribute to the whole fetch of the file
  • The speed test should compare to fetching the file using HTTP on the same network
@goga-m
Copy link
Contributor

goga-m commented Nov 27, 2017

I appreciate @diasdavid, thanks! Let's make it dance :)

Good points, especially peer related information on files is something that will really give a better view of what's going on in the background. And speed comparison with HTTP seems essential too. Another thing that comes to my mind now is:

  • We could also add speed test for put command as well
    And as @victorbjelkholm pointed here (benchmark.js.ipfs.io ipfs/js-ipfs#946 (comment))
  • Be able to run tests multiple times (would be good to test any browser crash issues)
  • Pining/unpinning files so that the user will be able to clear locally cached files (that could also include to create a view of lists of locally added files and the relative actions: Add file, remove/unping)

Sign up for free to subscribe to this conversation on GitHub. Already have an account? Sign in.
Labels
Projects
None yet
Development

No branches or pull requests

2 participants