YaCy – A Decentralized Peer-to-Peer Search Engine: Installation and Quick Start Guide

[Kulle]

YaCy is a free, decentralized search engine based on a peer-to-peer (P2P) network. Unlike traditional search engines, there is no central server that processes all search queries or controls the entire search index. Instead, each participant runs their own "peer", which crawls websites, indexes their content, and shares the collected data with other peers.

Documentation:
https://yacy.net/docs/

Download:
https://yacy.net/download_installation/#package-linux

Extract and start:

Code: Select all

tar xfz yacy<VERSION>.tar.gz
cd yacy<VERSION>
./startYACY.sh

Then open the following URL in a web browser:
http://localhost:8090/

The YaCy web interface should then appear.

Note: Java is required.

If Java is not installed, you can download an OpenJDK module here:
https://www.mediafire.com/file/0hk9sy0w ... 0.xzm/file

To ensure that YaCy can find Java, you may need to run the following command:

Code: Select all

export PATH=/usr/lib64/java/bin:/usr/local/bin:/usr/bin:/bin:/usr/games

[vinnie]

Thanks for the guide; I'll probably give it a try.

But tell me something: if a site isn't indexed, how does it index it?
Does the program run web crawlers locally?

Edit 1:
1st problem, If i change the theme, it's is effective only to the apperace configuration page a not outside. Can you change the theme?

[Kulle]

Hi vinnie,

1) Choose operating mode
Best: Peer Mode
Provides real results from the YaCy network
2) Adjust memory
Increase Java Heap Memory (1–4 GB)
Enable RAM Index if possible
3) Start crawling (basic index)
Begin with a few seed URLs
Keep crawl depth small (1–3)
Don’t crawl too much at once
4) Check peer connections
Make sure peer is connected
Enable index sharing
Otherwise results are very limited
5) Set crawl filters
Index only useful content (HTML, PDF)
Exclude login pages, session URLs, and junk
6) Configure index
Combine local + global index
Enable spam filters if available
7) Network / firewall
Open port 8090 (Web interface)
Check firewall/NAT if peers don’t connect
8) Enable updates
Turn on auto-update
Refresh peer list regularly
9) Set realistic expectations
Good results appear only after hours or days of crawling

[dreadbird]

This reminds me of helium a product my friend recommended that uses your internet bandwidth. they pay you monthly based on the bandwidth they used. something you install a wireless access point for it to connect to the network.

How does it find sites that arent indexed by someone else? does it use duckduckgo or something to get the site lists?