NNTP Version: 2.3.2.1
Platform : SunOS news1 5.6 Generic_105182-04 i86pc
Sorry if the following problems have already been discussed/solved, but
maybe they are new:
I had been having seg fault problems with NNTPcache 2.3.2.1 under Solaris
2.6/x86 with the nocem daemon. I tracked down the guilty line causing
the seg faults:
line 239 of nocem.c causes the seg fault:
logd (("article %d yeilded %d new msgsid's and %d duplicates (already
classified msgid's)", artno, st->msgid_good - st2.msgid_good,
st->msgid_dup - st2.msgid_dup));
If I comment this line out then it no longer seg faults. I'm guessing
logd() has problems of some sort?
Also on line 366 of nocem.c:
return TRUE;
st->last_scan = time (NULL);
Shouldn't 'st->last_scan = time (NULL);' be before the 'return TRUE;'?
After fixing these two problems along with applying several patches from
this list (fixing the race conditions and article numbering problems) I've
found NNTPcache 2.3.2.1 to be much more reliable. However, I still
recieve bogus stats information and seg faults when trying to view it. Oh
well, I'm on the right track... =)
Hope this information helps!
---
Mike Wade * Senior System Administrator * Chattanooga Data Connection, Inc.
mwade@cdc.net * http://www.cdc.net/~mwade/ * 423-266-3369 x223
