🚀 go-pugleaf

In news.software.b, article <1gv12mINNcpm@grasp1.univ-lyon1.fr>,
  Christophe.Wolfhugel@grasp.insa-lyon.fr (Christophe Wolfhugel) writes:
> 
> A fully runninf "innd" needs a few thousands of pages. On my site
> (not yet a full feed), the size seems to stabilize at around 2500 pages
> which is 10 Megs.  Being on a 16 Megs system, it mostly gets swapped
> out.  On incoming NNTP connections, I have seen around 1000 pages being
> loaded.  I also run INN on a 8 Megs Sun 3. Just fine.  My history is
> relatively small, around 7 Megs.
> 
Sparc, 15 outgoing feeds, history is 28 MB (the .pag file, which is what INN
keeps in core, is about 3 MB):

PID TT STAT  TIME SL RE PAGEIN SIZE  RSS   LIM %CPU %MEM COMMAND
207 ?  S     2:56 19 20   7014 6784  620    xx  0.0  1.0 innd

This is after an uptime on 10 hours and with two incoming feeds (which are
nntplink-run and probably idle right now).

> Maybe could Rich post UUNET's memory pages occupation of innd regarding
> to their history's size which is just in "another dimension" compared to
> my 7 Megs.
> 
I think they now have a 14-day cutoff, which means that keeping more than 
three weeks of history is wasting disk space and CPU time. (The number
of "late" articles which would have to be NNTP-transmitted to be recognized
as too old is too small to affect things.)

> > c)      does fairly important'' mean that INN will crash and
> >        burn if it can't malloc  more than 2.5 MB of VM (e.g.)?
> 
> Did anyone try with so little ?
> 
It will crash and burn but if anybody runs a Usenet site with less than 16 MB
of swap configured, it's their own fault if things don't work.

> INN is doing in one process the job of msgidd + nntpd + relaynews, with
> no temporary files, less input-outputs and faster IO operations.
> 
But note that INN keeps everything in memory (this includes the full text of
every incoming article) and writes out everything in one fell swoop (that's
what writev() is there for, folks ;-). Depending on the file system code in
your kernel, this _may_ cause more system load _if_ other programs can't 
access the file system while INN writes out a new article, or an update of
history.pag.
Fortunately, most systems do this correctly.

-- 
If I were two-faced, would I be wearing this one?
	-- Abraham Lincoln (1809-1865)
-- 
Matthias Urlichs  --  urlichs@smurf.sub.org -- urlichs@smurf.ira.uka.de   /(o\
Humboldtstrasse 7 -- 7500 Karlsruhe 1 -- Germany  --  +49-721-9612521     \o)/