Re: cfservd segmentation faults

[Mark . Burgess]

THe problem will probably occur again as you get new hosts. AS long as the hosts
ghave already exchanged trusted keys, the don't need to do a reverse lookup
in dns, so the offending call is avoided.

Mark


On 27 Sep, Jeremy Dreese wrote:
> 
> Upgrading to v2.1.0b3 seems to have fixed the problem.  I'll admit that I'm a
> little curious as to why (if you don't mind taking the time to explain).
> 
> Thanks a bunch!!
> 
> Quoting Mark.Burgess@iu.hio.no:
>> 
>> This looks like the usual bug in the resolver libraries.
>> Somewhere in nsswitch.conf or equivalent, you might have
>> something that says you looks up hostnames in NIS or files
>> before dns. This tends to cause this problem. 
>> 
>> You can also try 2.1.0b3 which might solve the problem in
>> a different way...
>> 
>> M
>> 
>> On 27 Sep, Jeremy Dreese wrote:
>> > 
>> > I wasn't sure whether to post this here or the bug list but here it is...
>> > 
>> > I recently upgraded from version 2.0.4 to 2.0.8p1 and I've been receiving
>> > segmentation faults from cfservd on my cfengine server.  Debugging mode
>> > (--debug) didn't seem to provide helpful info because it seemed to seg
>> fault at
>> > different places each time (probably just different amounts of output
>> buffered
>> > before the seg fault?).  So I recompiled with debugging on (-g to compiler)
>> and
>> > had cfservd crash twice with similar info.  I don't know much about using
>> gdb,
>> > but from a previous post I saw that it was suggested to run "gdb -c core
>> > /path/to/cfservd" and then type "backtrace".  Here's what I received:
>> > 
>> > Core was generated by `/var/cfengine/bin/cfservd --debug'.
>> > Program terminated with signal 11, Segmentation fault.
>> > #0  0x40095bc2 in ?? ()
>> > (gdb) backtrace
>> > #0  0x40095bc2 in ?? ()
>> > #1  0x4008c561 in ?? ()
>> > #2  0x4008c61e in ?? ()
>> > #3  0x40093ee5 in ?? ()
>> > #4  0x4001e323 in ?? ()
>> > #5  0x4001ecd0 in ?? ()
>> > #6  0x08170020 in gethostbyname2_r ()
>> > #7  0x0816a5c0 in gaih_inet ()
>> > #8  0x0816aba2 in getaddrinfo ()
>> > #9  0x0804bec2 in VerifyConnection (conn=0x40259700, 
>> >     buf=0xbadfea62 "134.82.133.1 alcor.eg.bucknell.edu root 0")
>> >     at cfservd.c:1792
>> > #10 0x0804a581 in BusyWithConnection (conn=0x40259700) at cfservd.c:1212
>> > #11 0x0804a1c3 in HandleConnection (conn=0x40259700) at cfservd.c:1112
>> > #12 0x08136b2b in pthread_start_thread ()
>> > 
>> > I'm wondering if any of this is load related.  I have around 100-110 hosts
>> > connecting to one server with a splay time of 30 minutes.
>> > 
>> > Let me know if there's anything else I can provide.  Thanks in advance for
>> the help!
>> > 
>> > 
>> > 
>> > _______________________________________________
>> > Help-cfengine mailing list
>> > Help-cfengine@gnu.org
>> > http://mail.gnu.org/mailman/listinfo/help-cfengine
>> 
>> 
>> 
>> ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
>> Work: +47 22453272            Email:  Mark.Burgess@iu.hio.no
>> Fax : +47 22453205            WWW  :  http://www.iu.hio.no/~mark
>> ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
>> 
>> 
> 
> 
> -----------------------------------------
> Jeremy Dreese
> Engineering Computing Systems Integrator
> College of Engineering
> Bucknell University
> voice: (570) 577-3714
> fax: (570) 577-3579
> email: jdreese@bucknell.edu



~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
Work: +47 22453272            Email:  Mark.Burgess@iu.hio.no
Fax : +47 22453205            WWW  :  http://www.iu.hio.no/~mark
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~