RE: [leafnode-list] Re: Question regarding Fetchnews

["Bill Gross" <bill@xxxxxxxxxxxxxxxx>]

I think the server disconnects by design, because the server always "goes
away" after Fetchnews has been connected for about 21600 seconds.

Following are excerpts from my logs with debugmode set at 4045:

Apr  1 01:00:57 gateway fetchnews[9199]: store: read 83 bytes:
"X-Received-Date: Mon, 31 Mar 2003 14:02:49 PST
(newsmaster1.news.pas.earthlink.net)", to go: -1
Apr  1 01:00:57 gateway fetchnews[9199]: store: read 0 bytes: "", to go: -1
Apr  1 01:01:35 gateway fetchnews[9199]: received article, sent ARTICLE
1092159 command, in pipe: 15
Apr  1 01:01:35 gateway fetchnews[9199]: Server went away when it should
send an article.
Apr  1 01:01:35 gateway fetchnews[9199]: alt.binaries.gdead: 1069 articles
fetched (to 44970), 0 killed
Apr  1 01:01:35 gateway fetchnews[9199]: found 0 articles in out.going.
Apr  1 01:01:35 gateway fetchnews[9199]: news.earthlink.net: 0 articles
posted
Apr  1 01:01:35 gateway fetchnews[9199]: mergesort(base=0x40173008,
nmemb=30422, size=28, compar=0x804ef20)
Apr  1 01:01:35 gateway fetchnews[9199]: sort took 30421 comparisons, called
from activutil.c:283
Apr  1 01:01:36 gateway fetchnews[9199]: wrote groupinfo with 30422 lines.
Apr  1 01:01:36 gateway fetchnews[9199]: fetchnews: 1505 articles and 0
headers fetched, 465 killed, 0 posted, in 21636 seconds

Apr  1 07:16:04 gateway fetchnews[9499]: store: read 83 bytes:
"X-Received-Date: Mon, 31 Mar 2003 15:15:36 PST
(newsmaster1.news.pas.earthlink.net)", to go: -1
Apr  1 07:16:04 gateway fetchnews[9499]: store: read 0 bytes: "", to go: -1
Apr  1 07:16:21 gateway fetchnews[9499]: received article, sent ARTICLE
12472489 command, in pipe: 15
Apr  1 07:16:21 gateway fetchnews[9499]: Server went away when it should
send an article.
Apr  1 07:16:21 gateway fetchnews[9499]: alt.binaries.vcd: 620 articles
fetched (to 12566), 0 killed
Apr  1 07:16:21 gateway fetchnews[9499]: found 0 articles in out.going.
Apr  1 07:16:21 gateway fetchnews[9499]: news.earthlink.net: 0 articles
posted
Apr  1 07:16:21 gateway fetchnews[9499]: mergesort(base=0x40173008,
nmemb=30422, size=28, compar=0x804ef20)
Apr  1 07:16:21 gateway fetchnews[9499]: sort took 30421 comparisons, called
from activutil.c:283
Apr  1 07:16:22 gateway fetchnews[9499]: wrote groupinfo with 30422 lines.
Apr  1 07:16:22 gateway fetchnews[9499]: fetchnews: 1326 articles and 0
headers fetched, 944 killed, 0 posted, in 21622 seconds

Apr  1 13:30:57 gateway fetchnews[9714]: store: read 83 bytes:
"X-Received-Date: Mon, 31 Mar 2003 10:46:24 PST
(newsmaster1.news.pas.earthlink.net)", to go: -1
Apr  1 13:30:57 gateway fetchnews[9714]: store: read 0 bytes: "", to go: -1
Apr  1 13:31:19 gateway fetchnews[9714]: received article, sent ARTICLE
1912202 command, in pipe: 15
Apr  1 13:31:19 gateway fetchnews[9714]: Server went away when it should
send an article.
Apr  1 13:31:19 gateway fetchnews[9714]: alt.binaries.southpark: 99 articles
fetched (to 1702), 0 killed
Apr  1 13:31:19 gateway fetchnews[9714]: found 0 articles in out.going.
Apr  1 13:31:19 gateway fetchnews[9714]: news.earthlink.net: 0 articles
posted
Apr  1 13:31:20 gateway fetchnews[9714]: mergesort(base=0x40173008,
nmemb=30422, size=28, compar=0x804ef20)
Apr  1 13:31:20 gateway fetchnews[9714]: sort took 30421 comparisons, called
from activutil.c:283
Apr  1 13:31:20 gateway fetchnews[9714]: wrote groupinfo with 30422 lines.
Apr  1 13:31:20 gateway fetchnews[9714]: fetchnews: 1950 articles and 0
headers fetched, 398 killed, 0 posted, in 21620 seconds

-----Original Message-----
From: Matthias Andree [mailto:ma@xxxxxxxxxxxxxxxxxxxxxxxxxxxx] 
Sent: Saturday, March 29, 2003 3:04 PM
To: leafnode-list@xxxxxxxxxxxxxxxxxxxxxxxxxxxx
Subject: Re: [leafnode-list] Re: Question regarding Fetchnews

"Bill Gross" <bill@xxxxxxxxxxxxxxxx> writes:

> 	My question really is, when the server "goes away" can Fetchnews be
> configured to reconnect to the same group it had been working on and
> continue down the list of interesting groups? I understand from Matthias,
> response that it's not currently possible, but do you think the code could
> be modified to do this? I'm running leafnode-2.0.0.alpha20030318a.

It might be doable, because after the disconnection, with a fresh
connection, the state would be known (modulo upstream server
bugs). However, as the reason for the disconnection is unclear, we
should figure that out first. Can you see from the debug log how much
time has passed since the previous command until the disconnect?
I. e. is that a timeout? If so, upstream server and fetchnews might get
out of sync. If not, but if the disconnect always happens at the exact
same place or time, the upstream server might be buggy.

-- 
Matthias Andree
leafnode-1 download: http://sourceforge.net/projects/leafnode/
leafnode-1 docs/new: http://mandree.home.pages.de/leafnode/
leafnode-2 homepage: http://mandree.home.pages.de/leafnode/beta/

-- 
leafnode-list@xxxxxxxxxxxxxxxxxxxxxxxxxxxx -- mailing list for leafnode
To unsubscribe, send mail with "unsubscribe" in the subject to the list

-- 
leafnode-list@xxxxxxxxxxxxxxxxxxxxxxxxxxxx -- mailing list for leafnode
To unsubscribe, send mail with "unsubscribe" in the subject to the list