How does INN store articles?
The method that INN uses to store Usenet messages is practically unchanged from the first days of Usenet. Each message is stored into its own file, in a subdirectory based on the newsgroup name. The name of the file is determined sequentially, on a per-group and per-server basis. (Article numbers do not match across servers.)
Sometimes an article is posted to more than one newsgroup. the preferred method is to create multiple links to the file. If this is not supported, then a copy is made and stored for each newsgroup.
The time the article was received, the message ID, where it was stored, among other things, are stored in a "history database." This database is used to remove expired articles, and to locate articles based on message ID.
There are various alternative methods, each with efficiencies and drawbacks. See Usenet RKT For Developers for more information.
Up To Usenet RKT: How does a newsserver work?
RKT Rapid-Links:[Search] [RKT Tips] Path: / Usenet RKT / For Providers / Getting Started / 0224.htm
You are reading from the Usenet RKT
Comments? [DocID: USERKT/INN/0224.htm]
Copyright 1997-1998, Forrest J. Cavalier III, Mib Software, INN customization and consulting