http://bugzilla.spamassassin.org/show_bug.cgi?id=3828
------- Additional Comments From [EMAIL PROTECTED] 2004-09-28 09:21 ------- err.. now i'm going to have to wait until it happens again. seems to be popping up about every 2 hours or so. i've come up with a lot more info in the meantime, maybe this will be more helpful? spamd1 has gone away 3 times this morning.. everytime it does (and comes back usually 30+ minutes later), you see the 'bad protocol' dbg message in the log. from 9:11 to 9:27 GMT it stopped. @4000000041592ae93729be04 2004-09-28 09:11:59 [20696] i: connection from p35n0.ruraltel.net [24.225.0.35] at port 60169 @4000000041592e7105aea894 2004-09-28 09:27:03 [9152] i: bad protocol: header error: (closed before headers) from 9:27 to 10:48 GMT it stopped. @4000000041592e7f0cdbaa04 2004-09-28 09:27:17 [9152] i: connection from p36n0.ruraltel.net [24.225.0.36] at port 50510 @40000000415941a135605c2c 2004-09-28 10:48:55 [9152] i: bad protocol: header error: (closed before headers) from 10:49 to 12:26 GMT it stopped. @40000000415941ac2dd64e6c 2004-09-28 10:49:06 [9152] i: connection from p35n0.ruraltel.net [24.225.0.35] at port 59890 @40000000415958941293ed94 2004-09-28 12:26:50 [20696] i: bad protocol: header error: (closed before headers) when it was down this last time, i decided to kill -9 15515 a.k.a the last 'spamd child' on my ps listing. once i did that, approx 130 email's were processed by the spamd children. The weird thing is, this box was completely out of the DNS and would not be accepting 'new' connections. all those message had to be sitting in the tcp sockets waiting to be processed, and once the childs became available, they flooded through? also saw a warn() and a dbg about a dead tcp socket, after i killed that one spamd child. @40000000415981e41bfd1f7c Use of uninitialized value in subroutine entry at /usr/lib/perl5/5.6.1/i386-linux/Socket.pm line 312. @40000000415981e41c00924c 2004-09-28 15:23:06 [24774] i: error: Bad arg length for Socket::unpack_sockaddr_in, length is 0, should be 16 at /usr/lib/perl5/5.6.1/i386-linux/Socket.pm line 312._ Transport endpoint is not connected, continuing so the 'bad protocol: header error' seems to be the constant factor here of when it comes back online by itself. question is what happens to the previous 'connection from' that causes everything else to stop... from spamd4 yesterday.. i: result: Y 5 - BAYES_20,DCC_CHECK,HTML_90_100,HTML_IMAGE_RATIO_04,HTML_MESSAGE,MIME_BOUND_NEXTP ART,OPTING_OUT,URIBL_SBL,URIBL_WS_SURBL scantime=0.3,size=8798,mid=<[EMAIL PROTECTED]>,bayes=0.098527 1200051288,autolearn=no i: connection from p34n0.ruraltel.net [24.225.0.34] at port 47706 i: bad protocol: header error: (closed before headers) i: connection from p35n0.ruraltel.net [24.225.0.35] at port 38263 on spamd3 yesterday i: result: Y 11 - BAYES_50,URIBL_OB_SURBL,URIBL_SBL,URIBL_SC_SURBL,URIBL_WS_SURBL scantime=0.2,size=2940,mid=<[EMAIL PROTECTED]>, bayes=0.573393109179598,autolearn=no i: connection from p35n0.ruraltel.net [24.225.0.35] at port 37916 i: bad protocol: header error: (closed before headers) i: connection from p35n0.ruraltel.net [24.225.0.35] at port 38257 from spamd4 on the 25th.. i: result: . 0 - AWL,BAYES_00,HTML_MESSAGE,HTML_NONELEMENT_00_10,HTML_TEXT_AFTER_BODY,HTML_TEXT_A FTER_HTML,MIME_HTML_ONLY,TO_ADDRESS_EQ_REAL,URIBL_SBL scantime=1.3,size=43426,mid=<[EMAIL PROTECTED] mail.com>,bayes=0,autolearn=no i: connection from p36n0.ruraltel.net [24.225.0.36] at port 59415 i: bad protocol: header error: (closed before headers) i: connection from p35n0.ruraltel.net [24.225.0.35] at port 45541 is anyone else seeing this 'header error'? ------- You are receiving this mail because: ------- You are the assignee for the bug, or are watching the assignee.
