Bugs item #3526603, was opened at 2012-05-14 08:20
Message generated for change (Comment added) made by hemanthreddy
You can respond by visiting:
https://sourceforge.net/tracker/?func=detail&atid=532251&aid=3526603&group_id=71730
Please note that this message will contain a full copy of the comment thread,
including the initial issue submission, for this request,
not just the latest update.
>Category: HP c-Class Plugin
Group: 3.2.0
>Status: Closed
>Resolution: Fixed
Priority: 5
Private: No
Submitted By: Tim Crawford - Tekelec (tcrawfo3)
Assigned to: dr_mohan (dr_mohan)
Summary: openhpid core with SIGTERM
Initial Comment:
We are using a variant of Centos 6.2 with Openhpi-3.0.0 libraries (64-bit).
We also include a bug fix patch for 3025938.
openHPId core's when we issue a SIGTERM to the process.
We did not see this issue with openhpid 2.14.1 we just started using 3.0.0.
We run openhpid to communicate with the HP-C class Onboard Administrator
Here is the backtrace:
Thread 1 (Thread 0x7fcf6e7377e0 (LWP 3766)):
#0 0x00000033e6275325 in malloc_consolidate () from /lib64/libc.so.6
No symbol table info available.
#1 0x00000033e62783b2 in _int_malloc () from /lib64/libc.so.6
No symbol table info available.
#2 0x00000033e62797d1 in malloc () from /lib64/libc.so.6
No symbol table info available.
#3 0x00000033e865d95e in CRYPTO_malloc () from /usr/lib64/libcrypto.so.10
No symbol table info available.
#4 0x00000033ec6231da in ssl3_new () from /usr/lib64/libssl.so.10
No symbol table info available.
#5 0x00000033ec62b279 in tls1_new () from /usr/lib64/libssl.so.10
No symbol table info available.
#6 0x00000033ec638c91 in SSL_new () from /usr/lib64/libssl.so.10
No symbol table info available.
#7 0x00000033ec643106 in BIO_new_ssl () from /usr/lib64/libssl.so.10
No symbol table info available.
#8 0x00000033ec64319d in BIO_new_ssl_connect () from /usr/lib64/libssl.so.10
No symbol table info available.
#9 0x00000033e8e01d94 in oh_ssl_connect (hostname=0x1db6968 "10.240.34.8:443", ctx=0x1db74e0, timeout=20) at oh_ssl.c:464
bio = 0x7fff5175d510
ssl = 0x42bb92
readfds = {fds_bits = {140734560059952, 140528885606200, 401, 1, 31148816, 4387352, 593, 1, 6514032, 222933988636, 140528888018912, 888088677683186931, 1366676752, 140528885351211, 140734560060096, 140528885351211}}
writefds = {fds_bits = {4294967296, 888088677683186931, 32613312, 140528885351211, 140734560060128, 888088677683186931, 140734560064784, 140528885351211, 140734560062272, 4223984, 140734560064784, 0, 0, 222933860323, 206158430232, 140734560060144}}
tv = {tv_sec = 4294967296, tv_usec = 1}
fd = 1366672144
err = 32767
#10 0x00007fcf6e4e8bcb in soap_message (connection=0x1db6950, request=0x1db6ac8 "<?xml version=\"1.0\"?>\n<SOAP-ENV:Envelope xmlns:SOAP-ENV=\"http://www.w3.org/2003/05/soap-envelope\" xmlns:xsi=\"http://www.w3.org/2001/XMLSchema-instance\" xmlns:xsd=\"http://www.w3.org/2001/XMLSchema\" xml"..., doc=0x1db6ac0) at oa_soap_callsupport.c:640
nbytes = 1366676240
ret = 32767
header = 0x0
response = "\030\000\000\000\060\000\000\000PËuQÿ\177\000\000\220ÊuQÿ156ÿÿÿÿ\000\000\000\000\030\000\000\000\060\000\000\000pËuQÿ\177\000\000°ÊuQÿ\177\000\000ÿÿÿÿ\000\000\000\000\030\000\000\000\060\000\000\000\220ËuQÿ\177\000\000ÐÊuQÿ\177\000\000\000'\000\000\000\000\000\000P×\006\002\000\000\000\000PÊuQÿ\177\000\000\003\000\000\000\000\000\000\000`ÊuQÿ\177\000\000\237\071\065æ3\000\000\000y>$æ3", '\000' <repeats 11 times>, "\001", '\000' <repeats 15 times>, "\001\000\000\000\000\000\000\000°ÊuQÿ\177\000\000| ÊuQÿ\177\000\000@..."...
parse = 0x4073f0
#11 0x00007fcf6e4e9a72 in soap_call (connection=0x1db6950) at oa_soap_callsupport.c:1079
session_pos = 0x1db6cd0 "f5f2752c9dfb070b</hpoa:oaSessionKey>\n</hpoa:HpOaSessionKeyToken>\n</wsse:Security>\n</SOAP-ENV:Header>\n<SOAP-ENV:Body>\n<hpoa:userLogOut/>\n</SOAP-ENV:Body>\n</SOAP-ENV:Envelope>\n"
i = 0
err = 51
#12 0x00007fcf6e4e96a1 in soap_logout (connection=0x1db6950) at oa_soap_callsupport.c:901
No locals.
#13 0x00007fcf6e4e8a1f in soap_close (connection=0x1db6950) at oa_soap_callsupport.c:570
No locals.
#14 0x00007fcf6e4aca4e in oa_soap_close (oh_handler=0x1db37c0) at oa_soap.c:384
handler = 0x1db37c0
oa_handler = 0x1db4600
#15 0x0000000000414954 in oh_close_handlers ()
No symbol table info available.
#16 0x0000000000413988 in oh_finit ()
No symbol table info available.
#17 0x0000000000412263 in main ()
No symbol table info available.
----------------------------------------------------------------------
>Comment By: Hemantha Beecherla (hemanthreddy)
Date: 2012-07-18 11:20
Message:
Fixed in Rev.7481
----------------------------------------------------------------------
Comment By: Hemantha Beecherla (hemanthreddy)
Date: 2012-07-18 02:04
Message:
Moving the group from 3.0.0 to 3.2.0
----------------------------------------------------------------------
Comment By: Hemantha Beecherla (hemanthreddy)
Date: 2012-06-01 10:08
Message:
Yes, this issue will be resolved once the fix is available for openssl
issue.
----------------------------------------------------------------------
Comment By: Tim Crawford - Tekelec (tcrawfo3)
Date: 2012-06-01 10:00
Message:
That is fine with me. I am assuming the fix to openssl would resolve the
issue as seen in openhpi.
----------------------------------------------------------------------
Comment By: Hemantha Beecherla (hemanthreddy)
Date: 2012-06-01 09:58
Message:
Hi Tim,
I have raised an issue in CentOS Bug Tracker on openssl.
0005751: Segfault obesrved in openssl-1.0.0-20
(http://bugs.centos.org/view.php?id=5751)
Is it ok for you if this issue is closed here?.
----------------------------------------------------------------------
Comment By: Tim Crawford - Tekelec (tcrawfo3)
Date: 2012-05-25 09:46
Message:
Hey, thanks for the feedback. We were using Centos 5.7 until recently when
we jumped to 6.2, so sounds like this is an openssl issue?
I am ok with whatever course of action you deem appropriate. If need be I
can just issue a SIGKILL instead of SIGTERM.
I am not able to use the openhpid stop script at this time since we are
actually running multiple openhpid. One per HP Enclosure.
----------------------------------------------------------------------
Comment By: Hemantha Beecherla (hemanthreddy)
Date: 2012-05-25 09:04
Message:
Hi Tim,
I have found that this issue is not reproduced using the script
/usr/local/etc/init.d/openhpid stop.
Script is using killproc command to kill the openhpid daemon as below,
killproc /usr/local/sbin/openhpid
Killproc will send the SIGTERM by default if nothing specified, if the
process doesn’t kill in 5 second, SIGKILL signal will be sent by
killproc. But the openhpid daemon is taking more than 15 seconds.
Obervation on other OS:
--> In Rhel5.3 OS this issue is not reproduced.
# cat /etc/issue
Red Hat Enterprise Linux Server release 5.3 (Tikanga) Kernel \r on an \m #
rpm -qa | grep openssl
openssl-devel-0.9.8e-7.el5
openssl-0.9.8e-7.el5
--> In CentOS5.6 OS this issue is not reproduced.
[root@... ~]# cat /etc/issue
CentOS release 5.6 (Final)
Kernel \r on an \m
# rpm -qa | grep openssl
openssl-devel-0.9.8e-12.el5_5.7
openssl097a-0.9.7a-9.el5_4.2
openssl-0.9.8e-12.el5_5.7
--> In CentOS6.2 This issue is reproduced only without script(ie with #kill
PID).
# cat /etc/issue
CentOS release 6.2 (Final)
Kernel \r on an \m
# rpm -qa | grep openssl
openssl-devel-1.0.0-20.el6.i686
openssl-1.0.0-20.el6.i686
With above observation, we will file a bug on the distributor or on
openssl-1.0.0-20.
Let me know your thoughts.
Hemantha Reedy
----------------------------------------------------------------------
Comment By: dr_mohan (dr_mohan)
Date: 2012-05-18 13:44
Message:
Tim,
Anton was mentioning the difference between the releases.
We are working on it. Looks like we need to dig deeper and look at ssl
also.
Mohan
----------------------------------------------------------------------
Comment By: Anton Pak (avpak)
Date: 2012-05-14 09:47
Message:
Just quick note with information about the difference.
HP guys will say more, I guess.
----------------------------------------------------------------------
Comment By: Tim Crawford - Tekelec (tcrawfo3)
Date: 2012-05-14 09:45
Message:
So is this something I need to address in our code? Or were you just
letting me know the difference between releases?
----------------------------------------------------------------------
Comment By: Anton Pak (avpak)
Date: 2012-05-14 08:26
Message:
Brief comment:
In 2.14 SIGTERM and SIGINT were not handled.
So openhpid was just terminated.
Now the signals are handled for graceful shutdown.
And the finalization code is called for the plug-in.
Probably nobody called this code before.
----------------------------------------------------------------------
You can respond by visiting:
https://sourceforge.net/tracker/?func=detail&atid=532251&aid=3526603&group_id=71730
|