[bisq-network/bisq] Create metric to monitor network size (#3916)
Florian Reimair
notifications at github.com
Wed Jan 29 14:10:29 UTC 2020
spend some time on this and IMHO the proof of concept turned out even better than I expected.
# TL;DR
IMO pretty good results. As expected, they are not spot on, but they certainly are close enough for our purposes.
Will proceed on simulating a more dynamic network where nodes come and go.
# Setup
I created a simulation environment as follows:
![Screenshot from 2020-01-29 14-36-10](https://user-images.githubusercontent.com/1070734/73361640-a09e7e00-42a5-11ea-8005-cc49ddb5bcc4.png)
- *Bisq Simulator*: queries the API of the price node just as a full Bisq client does, except
- query every 2 seconds (instead of 60)
- only provide the Bisq version information (instread of bisq version + uid of the http client)
- *Sim Controller*: can spawn lots of *Bisq Simulators*
- *Bisq Pricenode*: run from master (removed BTCAverage though)
- *Scraper*: the Device Under Test and thus, the prototype of the whole metric.
- a shell script
- utilizes grep, cut, sed and tail
- runs every 21 seconds
# Results
I spawned 50 *Bisq Simulators*, ie. the exact number of Bisq clients up and running in the network is 50.
I let the *Scraper* do its job for approx. 3 minutes and recorded every result. This is what the metric got by scraping the price node logs:
![Screenshot from 2020-01-29 14-57-09](https://user-images.githubusercontent.com/1070734/73362961-160b4e00-42a8-11ea-9ac1-305ee7960a52.png)
Furthermore, the *Scraper* extracts the Bisq version spread. My *Sim Controller* randomly fired up these *Bisq Simulator* instances:
Version | Instance count
-- | --
v1.2.1 | 6
v1.2.2 | 14
v1.2.3 | 7
v1.2.4 | 9
v1.2.5 | 14
with these results:
![Screenshot from 2020-01-29 14-37-42](https://user-images.githubusercontent.com/1070734/73361606-8cf31780-42a5-11ea-8899-cdba863666ce.png)
--
You are receiving this because you are subscribed to this thread.
Reply to this email directly or view it on GitHub:
https://github.com/bisq-network/bisq/issues/3916#issuecomment-579772122
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.bisq.network/pipermail/bisq-github/attachments/20200129/feb5fbb3/attachment.html>
More information about the bisq-github
mailing list