[leafnode-list] Re: Fetchnews messages

["Matthias Andree" <matthias.andree@xxxxxx>]

Am 12.05.2009, 19:42 Uhr, schrieb Enrico Wiki  
<enrico.c.ml.address@xxxxxxxxx>:

> Hi,
> I am now using Leafnode-2.
> I've been testing Leafnode (1 and then 2) for four days now, on Ubuntu  
> 9.04.
> The first two days I had problems with my ADSL ISP,  so I had plenty of
> time-out disconnections with fetchnews. Like I said the culprit was my  
> ISP,
> but I had the feeling that Leafnode doesn't handle very well with a  
> slaggish
> connection, compared to some readers that could at least complete the  
> job.
> Anyway, yesterday my ISP got back to normal service and  Leafnode is  
> doing
> fine now.
> BTW, I like the fact that: 1. Leafnode uses little RAM and CPU (even when
> dealing with big newsgroups), and 2. Leafnode saves each article to an
> individual file, I think that's more reliable than storing a whole group
> into a big file.

Hi Enrico,

leafnode isn't made to find magic ways when underlying transports protocol  
or physical links fail. Leafnode relies on the operating system to handle  
TCP properly (most do this today for the common subset that leafnode  
uses); and when connections break, well, that's it. It's impossible to  
identify what is the cause - be it timeouts, "connection reset by peer",  
host/port unreachable messages, whatever - the assumption is that if you  
run fetchnews regularly (for instance, from a cron job), it will pick up  
where it left off the last time and things will sort themselves out. This  
usually works.

> When doing
> fetchnews -N my.favorite.newsgroup
> I got a good many 423 replies ( no such article in group)  from the news
> server, but I guess that's normal, just some canceled articles.

Perhaps - but usually a sign that the upstream server's database is  
corrupt/inconsistent, particularly the overview data doesn't match the  
available articles: the XOVER command offers articles that aren't  
available any more.

> I also  got a couple of "killed" articles, although I didn't set any
> filters. Maybe that depends on the
>   "store: duplicate article " ?

Likely, particularly if the fetchnews run was aborted some hard way.

HTH

-- 
Matthias Andree
-- 
_______________________________________________
leafnode-list mailing list
leafnode-list@xxxxxxxxxxxxxxxxxxxxxxxxxxxx
https://www.dt.e-technik.uni-dortmund.de/mailman/listinfo/leafnode-list
http://leafnode.sourceforge.net/