Indexing

Pages

The Admarus daemon automatically indexes all pages pinned on your IPFS node.

However, these pages can only be addressed by their cid using ipfs. Standalone pages are usually not ranked very well.

Complete Websites

Instead of standalone pages, Admarus allows adding whole websites. The daemon has the ability to keep track of DNS records for your domains and pin their data as they change.

The main benefit is that Admarus will know which page belongs to which domain. It will thus be able to format ipns urls from their domain names. This results in clearer search results and ranking benefits.

Use the argument dns-pins when launching the daemon to ask Admarus to pin a domain.

# Example for ipfs.tech
admarusd --dns-pins=ipfs.tech

You can also restrict the pin to a folder scope.

# Example to pin wikipedia articles only
admarusd --dns-pins=en.wikipedia-on-ipfs.org/wiki

There is no limit to how many domains you can pin.

Exotic data

Admarus is a search engine for the web. Only pages are indexed. Media content is not found directly, but from the pages embedding them.

Only HTML pages are currently supported, but support for MarkDown and PDF will be added in the future.

Any other container data needs to be transcribed to a supported format in order to be indexed.

Performance

Indexing takes place when you start the Admarus daemon. It will load all your pinned documents one by one, and store an index in RAM. The indexing process is slow and ressource-intensive. This is not a problem for regular users, indexing under a thousand documents. However, it must be considered if you are willing to contribute more to the network.

An on-disk index can be enabled, allowing Admarus to recover previously indexed documents when started. It is much more efficient, requiring only 5kB of RAM per document. To use this feature, please compile the daemon with the database-lmdb feature.