lwv27@LOCAL.UUCP (Larry W. Virden) (04/13/91)
Hi! We just brought up the latest (?) nntp on a Sparc server running SunOS 4.1 . We are seeing a very high percentage of the following types of messages: nntplog:Apr 13 04:10:23 srv01s4 nntpd[15777]: tut.cis.ohio-state.edu timeout nntplog:Apr 13 04:10:23 srv01s4 nntpd[15777]: tut.cis.ohio-state.edu timeout nntplog:Apr 13 05:26:12 srv01s4 nntpd[16554]: humpty unrecognized XHDR subject 5969 nntplog:Apr 13 06:06:42 srv01s4 nntpd[16950]: humpty unrecognized XHDR subject 32949 nntplog:Apr 13 07:05:14 srv01s4 nntpd[17315]: tut.cis.ohio-state.edu ihave_stats accepted 48 rejected 2 failed 0 nntplog:Apr 13 07:19:48 srv01s4 nntpd[17629]: humpty unrecognized XHDR subject 5989 nntplog:Apr 13 07:20:22 srv01s4 nntpd[17630]: humpty unrecognized XHDR subject 804 nntplog:Apr 13 07:41:32 srv01s4 nntpd[17704]: humpty unrecognized XHDR subject 12758 nntplog.o:Apr 12 05:37:58 srv01s4 nntpd[4929]: humpty unrecognized XHDR subject 40126 nntplog.o:Apr 12 06:48:01 srv01s4 nntpd[5410]: humpty unrecognized XHDR subject 5950 Note that the end result in these cases are folks not being able to read articles , only seeing headers of articles but receiving no article bodies, or rrn treating the entire group as not present, even though the files are really there. Anyone have any ideas about what is going on? Here are the vital stats: 1. Our NNTP software was updated to 201 srv01s4 NNTP[auth] server version 1.5.11 (10 February 1991) ready at Thu Apr 11 09:24:03 1991 (no posting). 2. It was moved from a MicroVax to a Sun 4/490 running 4.1. 3. Rn was updated to: @(#)$Header: rn.c,v 4.3.2.7 90/11/23 20:30:43 sob Exp $ Patch #: 50 Compiled under Sun's cc. 4. News delivery was switched to C News. The latest 4 patches have not as yet been applied as far as I am aware. Okay, the stage is set. When users are reading via rrn, they see three strange things: 1. rrn will get into a mode where it will only show headers and not message text or, in worse cases, indicate that no articles are available. When the user exits rrn with an x and returns, the group is fine. 2. rrn indicates a problem with nntp after a while . Often it is after a KILL file or an = type operation has occured. The error msg indicates that the nntp server was unable to find a particular group - even though it is there. One of the sample msgs is: rrn: server response to GROUP alt.text.dwb: Path: chemabs!tut.cis.ohio-state.edu!ucbvax!decwrl!pa.dec.com!bacchus!shepard In this case, rrn won't even go into the group. 3. At times, a user at the news selection level will press = to see what subjects are present. A list of subjects will appear. Then, when they press space, rrn indicates that it is skipping articles and will tell them that there are no articles remaining. While this would be normal if the articles are expired, what happens next is the strange part. When they press P to look at the previous message - they find that all of the articles which rrn just said it was skipping are still there - marked as having been seen. This happens quite frequently. Any ideas on what these might be? -- Larry W. Virden Business: UUCP: osu-cis!chemabs!lwv27 INET: lwv27%cas.BITNET@CUNYVM.CUNY.Edu Personal: 674 Falls Place, Reynoldsburg,OH 43068-1614 Pro-Line: lvirden@pro-tcc.cts.com America Online: lvirden CIS: [75046,606]