linux syslog problems (now linux-2.2.1+glibc-2.0.111)

[John Kennedy <jk@csuchico.edu>]

02/01/99 @ 05:39:19 AM (Monday)

  I've still chasing that problem.  I think the circumstances that
started the DNS/poll() problem are gone, but one of my secondary problems
that was taking out DHCP daemons is getting sendmail too.

  I've got a bunch of sendmails that are hung up like this:

    [sendmail pid=1117]
	#0  0x400d4672 in __libc_send ()
	#1  0x400d1131 in vsyslog ()
	#2  0x400d0e51 in syslog ()
	#3  0x805033b in sm_syslog ()
	#4  0x805d083 in logsender ()
	#5  0x805cdfa in eatheader ()
	#6  0x804d1a7 in collect ()
	#7  0x807911b in smtp ()
	#8  0x8060c87 in main ()
	#9  0x4003e8af in __libc_start_main ()

  I forgot to notice that I needed confOPTIMIZE now, so they weren't
compiled with `-g'.  This one has been a lot better about sticking around
and hits more daemons than sendmail more reliably.

  That is the child process:

	USER       PID %CPU %MEM  SIZE   RSS TTY STAT START   TIME COMMAND
	root      1116  0.0  1.3  2192  1328  ?  S    22:29   0:00 sendmail: server [::ffff:204.119.194.10] child wait          
	root      1117  0.0  1.4  2216  1368  ?  S    22:29   0:00 sendmail: WAA01117 [::ffff:204.119.194.10]: DATA             

  Note the time that it started -- they aren't going away fast.
The parent is just waiting on the child like you'd expect:

    [sendmail pid=1116]
	#0  0x400b4fd9 in __wait4 ()
	#1  0x4010fedc in __DTOR_END__ ()
	#2  0x804ec73 in waitfor ()
	#3  0x8079f17 in runinchild ()
	#4  0x8078a01 in smtp ()
	#5  0x8060c87 in main ()
	#6  0x4003e8af in __libc_start_main ()

  By comparison, here is DHCP:

	#0  0x400c6672 in __libc_send ()
	#1  0x400c3131 in vsyslog ()
	#2  0x400c2e51 in syslog ()
	#3  0x8059309 in note (fmt=0x805ab60 "BOOTREQUEST from %s via %s")
	    at errwarn.c:133
	#4  0x804d2f0 in bootp (packet=0xbfffe268) at bootp.c:70
	#5  0x80577c1 in do_packet (interface=0x8194290, packet=0xbfffeadc, len=300, 
	    from_port=17408, from={len = 4, 
	      iabuf = "\204ñ:\001", '\000' <repeats 11 times>}, hfrom=0xbffffaf4)
	    at options.c:607
	#6  0x805273f in got_one (l=0x81940d8) at dispatch.c:596
	#7  0x8052626 in dispatch () at dispatch.c:560
	#8  0x8049a86 in main (argc=1, argv=0xbffffd64, envp=0xbffffd6c) at dhcpd.c:278
	#9  0x400308af in __libc_start_main ()

  DHCP has been hung since ~Jan 31 12:30 trying to syslog that message.

  I'm going to retest these against glibc-2.0.112 by 10 AM or so.

  I really need to compile glibc with -g and recompile and I spent a
chunk of yesterday arranging things so it could be done.  That is an
extra 60MB of bloat on my box.  (:
								--- john