possibel to separate import header queue directory from rest

Technical support and discussion of Newsbin Version 6 series.

possibel to separate import header queue directory from rest

Postby spotter » Wed Aug 06, 2025 8:57 am

so I've gotten back to using newsbin pro after a decade away, and I realize that there's what I consider a major flaw in how it handles header import

1) without jumping through hoops of windows "bind" type mounting (though perhaps a bit easier if doing on linux in wine), one has to have the import and spool (_v6) directories on the same file system. This has an impact on ingest performance as doing heavy reads and writes at the same time to the same fs (especially if thte ingest falls behind the header download so no longer operating out of cache)

2) the header downloads are huge and ephemeral. i.e. one can easily download many many TBs of headers that are downloaded, written to disk and thrown away. For modern storage (i.e. SSDs) this is a terrible use of their life. These header files are perfectly suitable to spinning disks (they are basically just written and read sequentially) and using spinnign disk for them doesn't do anything to degrade the spinning disk, while using an SSD will use a substantial amount of its life. On the other hand, the spool directly is meant to be more persistent, and strongly benefits from the random access performance of SSDs (and just general performance), so one would always want it on an ssd if possible.

But i don't see a way to split them today.
spotter
Seasoned User
Seasoned User
 
Posts: 179
Joined: Tue Feb 12, 2002 12:00 pm

Registered Newsbin User since: 05/05/06

Return to V6 Technical Support

Who is online

Users browsing this forum: Bing [Bot] and 2 guests