github/gossipsub-hardening
1
1
Fork 0
mirror of https://github.com/libp2p/gossipsub-hardening.git synced 2026-01-08 21:48:00 -05:00
Code Issues Packages Projects Releases Wiki Activity

readme: fix file extension in examples

Browse Source
This commit is contained in:
Yusef Napora
2020-05-01 10:24:28 -04:00
committed by GitHub
parent 7f66dbea42
commit cee4f5749f
1 changed files with 7 additions and 7 deletions
Download Patch File Download Diff File Expand all files Collapse all files

14
README.md
View File

@@ -256,38 +256,38 @@ By default, it uses the S3 bucket `gossipsub-test-outputs` in `eu-central-1`, bu
To backup all the test outputs in `./output`:
```shell
./sync_outputs.sh store-all ./output
./sync_outputs.py store-all ./output
```
It will ignore the `failed` subdirectory automatically, but if you want to ignore more, you can pass in a flag:
```shell
./sync_outputs.sh store-all ./output --ignore some-dir-you-dont-want-to-store
./sync_outputs.py store-all ./output --ignore some-dir-you-dont-want-to-store
```
Alternatively, you can selectively store one or more test outputs with the `store` subcommand:
```shell
./sync_outputs.sh store ./output/pubsub-test-20200409-152658 ./output/pubsub-test-20200409-152983 # etc...
./sync_outputs.py store ./output/pubsub-test-20200409-152658 ./output/pubsub-test-20200409-152983 # etc...
```
You can also fetch test outputs from S3 to the local filesystem.
To fetch everything from the bucket into `./output`:
```shell
./sync_outputs.sh fetch-all ./output
./sync_outputs.py fetch-all ./output
```
Or, to fetch one or more tests from the bucket instead of everything:
```shell
./sync_outputs.sh fetch --dest=./output pubsub-test-20200409-152658
./sync_outputs.py fetch --dest=./output pubsub-test-20200409-152658
```
You can list all the top-level directories in the S3 bucket (so you know what to fetch) using the `list` command:
```shell
./sync_outputs.sh list
./sync_outputs.py list
```
## Code Overview
@@ -332,4 +332,4 @@ Because the full tracer output is quite large (several GB for a few minutes of t
we aggregate the trace events at runtime and spit out a json file with aggregated metrics at the end of the test.
We also capture a filtered subset of the original traces, containing only Publish, Deliver, Graft, and Prune events.
At the end of the test, we run [tracestat](https://github.com/libp2p/go-libp2p-pubsub-tracer/blob/master/cmd/tracestat/main.go)
on the filtered traces to calculate the latency distribution and get a summary of publish and deliver counts.
on the filtered traces to calculate the latency distribution and get a summary of publish and deliver counts.