Software Open Access

Netter: Probabilistic, Stateful Network Models

Zhang, Han; Zhang, Chi; Azevedo de Amorim, Arthur; Agarwal, Yuvraj; Fredrikson, Matt; Jia, Limin

Project Github Repo

The main Netter compiler is available at https://github.com/arthuraa/netter, while this artifact contains example code and models for case studies from the paper.

 

Extract Instruction

This artifact is designed to be executed with VMCAI 2021 Virtual Machine (https://doi.org/10.5281/zenodo.4017292).

One important assumption is that you extract the ZIP file content under `/home/vmcai2021`, since we have some implicit dependencies for local package locations. See README.md within the ZIP file for more details.

 

Description

This artifact contains the Netter compiler and models for all the cases discussed in the paper.

- Section 2: Example in the Overview
- Section 4.1: Warm up chain topology
- Section 4.2: MPLS
- Section 4.3: Load balancer
- Section 4.4: CoDef link-flooding defense

Moreover, it contains the Coq formalization.
The Coq development contains two main results in coq/Imp.v:
dead_store_elimP, dead_store_elim_optE, inline_run and inline_rew.
The first one says that the dead store elimination pass
"dead_store_elim" preserves the semantics of programs.  The second one
says that an optimized version of "dead_store_elim",
"dead_store_elim_opt", computes the same result. The last two say that
the inlining pass "inline" preserves the semantics of programs and of
computed rewards. These passes were translated by hand to the
"deadStoreElimOpt" and "inline" functions of
src/Netter/Compiler/Optimize.hs.

Please refer to the README.md file for additional details.

 

SHA256 checksum

9281e42609349a0da7d9094b0a5326af1a0e3d5d291892b0a5c3250ccf0cc88c  submission.zip

 

Files (1.2 GB)
Name Size
submission.sha256
md5:f11d2bbd0a5b841bd7334824d6bc4ae0
81 Bytes Download
submission.zip
md5:088739577940ad5289de9ffe2e395bc0
1.2 GB Download
131
21
views
downloads
All versions This version
Views 131131
Downloads 2121
Data volume 16.2 GB16.2 GB
Unique views 102102
Unique downloads 1616

Share

Cite as