From 7961b68852febc72c970b54bd7953459f9998e1a Mon Sep 17 00:00:00 2001
From: =?UTF-8?q?Rapha=C3=ABl=20Vinot?= <raphael@vinot.info>
Date: Mon, 12 Mar 2018 17:46:54 +0100
Subject: [PATCH] chg: Update readme

---
 README.md | 92 +++++++++++++++++++++++++++++++++++++++++++++++++++++++
 1 file changed, 92 insertions(+)

diff --git a/README.md b/README.md
index f57188b..fa1b94d 100644
--- a/README.md
+++ b/README.md
@@ -1,2 +1,94 @@
 # BGP-Ranking
 New version of BGP Ranking, complete rewrite in python3.6+ and an ARDB backend
+
+# Directory structure
+
+*Config files*: `listimport / modules_config / *.json`
+
+*Per-module parsers*: `listimport / parsers`
+
+*Libraries* : `listimport / libs`
+
+# Raw dataset directory structure
+
+## Files to import
+
+`<storage_directory> / <vendor> / <listname>`
+
+## Last modified date (if possible) and lock file
+
+`<storage_directory> / <vendor> / <listname> / meta`
+
+## Imported files less than 2 months old
+
+`<storage_directory> / <vendor> / <listname> / archive`
+
+## Imported files more than 2 months old
+
+`<storage_directory> / <vendor> / <listname> / archive / deep`
+
+# Databases
+
+## Intake (redis, port 6379)
+
+*Usage*: All the modules push their entries in this database.
+
+Creates the following hashes:
+
+```python
+UUID = {'ip': <ip>, 'source': <source>, 'datetime': <datetime>}
+```
+
+Creates a set `intake` for further processing containing all the UUIDs.
+
+
+## Pre-Insert (redis, port 6380)
+
+
+*Usage*: Make sure th IPs are global, validate input from the intake module.
+
+Pop UUIDs from `intake`, get the hashes with that key
+
+Creates the following hashes:
+
+```python
+UUID = {'ip': <ip>, 'source': <source>, 'datetime': <datetime>, 'date': <date>}
+```
+
+Creates a set `to_insert` for further processing containing all the UUIDs.
+
+Creates a set `for_ris_lookup` to lookup on the RIS database. Contains all the IPs.
+
+## Routing Information Service cache (redis, port 6381)
+
+*Usage*: Lookup IPs against the RIPE's RIS database
+
+Pop IPs from `for_ris_lookup`.
+
+Creates the following hashes:
+
+```python
+IP = {'asn': <asn>, 'prefix': <prefix>, 'description': <description>}
+```
+
+
+## Long term storage (ardb, port 16379)
+
+*Usage*: Stores the IPs with the required meta informations required for ranking.
+
+Pop UUIDs from `to_insert`, get the hashes with that key
+
+Use the IP from that hash to get the RIS informations.
+
+Creates the following sets:
+
+```python
+# All the sources, by day
+<YYYY-MM-DD>|sources = set([<source>, ...])
+# All the ASNs by source, by day
+<YYYY-MM-DD>|<source> -> set([<asn>, ...])
+# All the prefixes, by ASN, by source, by day
+<YYYY-MM-DD>|<source>|<asn> -> set([<prefix>, ...])
+# All the tuples (ip, datetime), by prefixes, by ASN, by source, by day
+<YYYY-MM-DD>|<source>|<asn>|<prefix> -> set([<ip>|<datetime>, ...])
+```