Google Groups no longer supports new Usenet posts or subscriptions. Historical content remains viewable.

Dismiss

[ace-bugs] ACE_Reactor: handle_events API returns -1

1 view

Skip to first unread message

prashant...@wipro.com

unread,

Aug 31, 2006, 9:50:50 AM8/31/06

to ace-...@cs.wustl.edu, rames...@wipro.com

Hi,

We are working for a Nortel product which is one of the sponsors listed. We are having a very critical issue with the ACE library which we are reporting. For one of our Element Management Systems for Telecom networks, we are using ACE as the communication framework. Your help will be appreciated in addressing this issue. Please let us know, if you need further clarification.

ACE VERSION: 5.1.17

HOST MACHINE and OPERATING SYSTEM:

# uname -a

HP-UX shivalik B.11.11 U 9000/785 2007660371 unlimited-user license

TARGET MACHINE and OPERATING SYSTEM, if different from HOST:

# uname -a

HP-UX shivalik B.11.11 U 9000/785 2007660371 unlimited-user license

COMPILER NAME AND VERSION (AND PATCHLEVEL):

what /usr/lib/libCsup.sl

/usr/lib/libCsup.sl:

HP aC++ B3910B A.03.70 Language Support Library

THE $ACE_ROOT/ace/config.h FILE [if you use a link to a platform-

specific file, simply state which one]:

/* -*- C++ -*- */

// $Id: config-hpux-10.x.h,v 4.5 1996/12/13 22:30:00 schmidt Exp $

// The following configuration file is designed to work for HP

// platforms running HP/UX 10.x. For using HP C++ compiler with

// templates you should have patch PHSS_6246.

#if !defined (ACE_CONFIG_H)

#define ACE_CONFIG_H

// Compiling for HPUX.

#if !defined (HPUX)

#define HPUX

#endif /* HPUX */

#define _HPUX_SOURCE

#include /**/ <sys/stdsyms.h>

#include /**/ <sched.h> /* pthread.h doesn't include this */

#define ACE_HAS_BROKEN_CONVERSIONS

// They forgot a const in the prototype of const_timewait...

#define ACE_LACKS_CONST_TIMESPEC_PTR

#define ACE_LACKS_SYSCALL

#define ACE_HAS_POSIX_TIME

// Compiler doesn't support static data member templates.

#define ACE_LACKS_STATIC_DATA_MEMBER_TEMPLATES

// Platform supports System V IPC (most versions of UNIX, but not Win32)

#define ACE_HAS_SYSV_IPC

// Platform supports recvmsg and sendmsg.

#define ACE_HAS_MSG

// Compiler/platform contains the <sys/syscall.h> file.

#define ACE_HAS_SYSCALL_H

// Fixes a problem with HP/UX not wrapping the mmap(2) header files

// with extern "C".

#define ACE_HAS_BROKEN_MMAP_H

// Prototypes for both signal() and struct sigaction are consistent.

#define ACE_HAS_CONSISTENT_SIGNAL_PROTOTYPES

// Compiler/platform has correctly prototyped header files.

#define ACE_HAS_CPLUSPLUS_HEADERS

// Header files lack t_errno for ACE_TLI.

#define ACE_LACKS_T_ERRNO

// Compiler/platform supports poll().

#define ACE_HAS_POLL

// Platform supports POSIX O_NONBLOCK semantics.

#define ACE_HAS_POSIX_NONBLOCK

// Compiler/platform defines the sig_atomic_t typedef

#define ACE_HAS_SIG_ATOMIC_T

//#define ACE_HAS_THREADS

//#define ACE_HAS_PTHREADS

//#define ACE_MT_SAFE

#define ACE_HAS_SIGINFO_T

//#define ACE_LACKS_PTHREAD_THR_SIGSETMASK

#define ACE_HAS_SETKIND_NP

#define ACE_LACKS_CONDATTR_PSHARED

#define ACE_LACKS_SI_ADDR

#define ACE_LACKS_SETDETACH /* new */

#define ACE_LACKS_KEYDELETE /* new */

//#define ACE_LACKS_THREAD_PROCESS_SCOPING

//#define ACE_LACKS_THREAD_STACK_ADDR

// Compiler/platform has thread-specific storage

//#define ACE_HAS_THREAD_SPECIFIC_STORAGE

// Compiler/platform supports struct strbuf.

#define ACE_HAS_STRBUF_T

// Compiler supports the ssize_t typedef.

#define ACE_HAS_SSIZE_T

// Defines the page size of the system.

#define ACE_PAGE_SIZE 4096

// Compiler/platform supports strerror ().

#define ACE_HAS_STRERROR

// SunOS 4 style prototype.

#define ACE_HAS_SUNOS4_GETTIMEOFDAY

// HP/UX has an undefined syscall for GETRUSAGE...

#define ACE_HAS_SYSCALL_GETRUSAGE

// Note, this only works if the flag is set above!

#define ACE_HAS_GETRUSAGE

// Platform uses int for select() rather than fd_set.

//#define ACE_SELECT_USES_INT

// Platform has the XLI version of ACE_TLI.

// #define ACE_HAS_XLI

// Platform supports ACE_TLI tiuser header.

#define ACE_HAS_TIUSER_H

// Platform provides ACE_TLI function prototypes.

//#define ACE_HAS_TLI_PROTOTYPES

// Platform supports ACE_TLI.

//#define ACE_HAS_TLI

// Turns on the tracing feature.

//#if !defined (ACE_NTRACE)

//#define ACE_NTRACE 0 // set to 1 to turn off

//#endif /* ACE_NTRACE */

#endif /* ACE_CONFIG_H */

THE $ACE_ROOT/include/makeinclude/platform_macros.GNU FILE [if you

use a link to a platform-specific file, simply state which one

(unless this isn't used in this case, e.g., with Microsoft Visual

C++)]:

Not available

CONTENTS OF $ACE_ROOT/bin/MakeProjectCreator/config/default.features

(used by MPC when you generate your own makefiles):

Not available

AREA/CLASS/EXAMPLE AFFECTED:

[What example failed? What module failed to compile?]

ACE_Reactor *pReactor

while (1) {

if( pReactor->handle_events()==-1)

{

syslog(LOG_ERR, "Error occurred in Reactor with errno %d , trying to recover..\n",errno);

}

Description: The handle_events() function in the above piece of code is returning -1.

DOES THE PROBLEM AFFECT:

COMPILATION?

LINKING?

On Unix systems, did you run make realclean first?

EXECUTION?

OTHER (please specify)?

[Please indicate whether ACE, your application, or both are affected.]

As ACE_Reactor ::handle_events() function returns -1 both ACE and our application are affected.

SYNOPSIS:

[Brief description of the problem]

Our application is a Element management system which is monitoring for events and sending requests to a Network Element. The applications functions fine for some time around 1 or 2 days and then at some point in time, the ACE_Reactor->handle_events() function starts returning -1 (error scenario) and it does not recover until the whole application is killed and restarted. We have tried to print the system error number which gets displayed as “9”.

DESCRIPTION:

[Detailed description of problem. Don't just say "<blah>

doesn't work, here's a fix," explain what your program does

to get to the <blah> state. ]

The application is monitoring for events and if there is one particular Network element to which the connection is not quite good (i.e. the requests made to the Network element from the EMS keep on timing out) then this problem is seen to occur. If the connectivity problem is rectified, the application doesn’t ever get in to the problem state. This is the only characterization that was possible. We tried relating it to the number of events or Network elements to check if there are any resources which the ACE framework is using are getting exhausted, but couldn’t characterize this.

REPEAT BY:

[What you did to get the error; include test program or session

transcript if at all possible. ]

If there is a network element to which the connection is not proper (say the connecting is lost), then the handle_events() API starts returning -1 and the application goes in to a “hung” state because f the while loop.

Since restart of the application solves the problem, we tried restarting the Reactor (by using .stop() and .start()) APIs), but this did not help. Only once you restart the application, the problem will not be seen (for some time again, may be 1 or 2 days).

SAMPLE FIX/WORKAROUND:

Restart of the application solves the problem but again for some time, hence not a solution. The workaround we used was to correct the connectivity problem to the Network element from the EMS after which the problem was not seen.

Regards,

Prashant Poulkar

- - - - - - - - - - - - - - - - - - - - - - - - - - - -

Wipro Technologies, Bangalore

Phone: +91 80 28520408 Extn 3269

- - - - - - - - - - - - - - - - - - - - - - - - - - - -

The information contained in this electronic message and any attachments to this message are intended for the exclusive use of the addressee(s) and may contain proprietary, confidential or privileged information. If you are not the intended recipient, you should not disseminate, distribute or copy this e-mail. Please notify the sender immediately and destroy all copies of this message and any attachments.

WARNING: Computer viruses can be transmitted via email. The recipient should check this email and any attachments for the presence of viruses. The company accepts no liability for any damage caused by any virus transmitted by this email.

www.wipro.com

prashant...@wipro.com

unread,

Aug 31, 2006, 9:53:00 AM8/31/06

to ace-...@mail.cse.wustl.edu, rames...@wipro.com

Sowayan, Abdullah (N-DUA)

unread,

Aug 31, 2006, 9:57:41 AM8/31/06

to prashant...@wipro.com, ace-...@cse.wustl.edu, rames...@wipro.com

You’re using a very ancient version of ACE. You should be upgrading to ACE 5.5.2. If that’s not a possibility, I think you’re better off contacting one of the commercial vendors of ACE to remedy your problem.

Thanks,

Abdul

Douglas C. Schmidt

unread,

Aug 31, 2006, 10:02:35 AM8/31/06

to prashant...@wipro.com, ace-...@cse.wustl.edu

Hi Prashant,

Wow, ACE 5.1.17 - that was 5 years ago! Please upgrade to ACE+TAO
x.5.2, which you can download from

http://deuce.doc.wustl.edu/Download.html

The DOC groups at Washington University, UC Irvine, and Vanderbilt
University only provide "best effort" support for non-sponsors for the
latest release, as described in

http://www.cs.wustl.edu/~schmidt/ACE_wrappers/docs/ACE-bug-process.html

Thus, if you need more "predictable" help for earlier versions of
ACE+TAO, I recommend that you check out

http://www.cs.wustl.edu/~schmidt/commercial-support.html

for a list of companies that will provide you with ACE+TAO commercial
support.

Thanks,

Doug

>We are working for a Nortel product which is one of the sponsors listed.
>We are having a very critical issue with the ACE library which we are
>reporting. For one of our Element Management Systems for Telecom
>networks, we are using ACE as the communication framework. Your help
>will be appreciated in addressing this issue. Please let us know, if you
>need further clarification.
>

>=0D
>
> ACE VERSION: 5.1.17
>
>=0D

>
> HOST MACHINE and OPERATING SYSTEM:
>
> # uname -a
>
> HP-UX shivalik B.11.11 U 9000/785 2007660371 unlimited-user license
>

>=0D

>
> TARGET MACHINE and OPERATING SYSTEM, if different from HOST:
>
> # uname -a
>
> HP-UX shivalik B.11.11 U 9000/785 2007660371 unlimited-user license
>

>=0D

>
> COMPILER NAME AND VERSION (AND PATCHLEVEL):
>

> what /usr/lib/libCsup.sl=0D

>
> /usr/lib/libCsup.sl:
>
> HP aC++ B3910B A.03.70 Language Support Library
>

>=0D

>
> THE $ACE_ROOT/ace/config.h FILE [if you use a link to a platform-
>
> specific file, simply state which one]:
>
> /* -*- C++ -*- */
>
>// $Id: config-hpux-10.x.h,v 4.5 1996/12/13 22:30:00 schmidt Exp $
>

>=0D

>
>// The following configuration file is designed to work for HP
>
>// platforms running HP/UX 10.x. For using HP C++ compiler with
>
>// templates you should have patch PHSS_6246.
>

>=0D

>
>#if !defined (ACE_CONFIG_H)
>
>#define ACE_CONFIG_H
>

>=0D

>
>// Compiling for HPUX.
>
>#if !defined (HPUX)
>
>#define HPUX
>
>#endif /* HPUX */
>

>=0D

>
>#define _HPUX_SOURCE
>
>#include /**/ <sys/stdsyms.h>
>
>#include /**/ <sched.h> /* pthread.h
>doesn't include this */
>

>=0D
>
>#define ACE_HAS_BROKEN_CONVERSIONS
>
>=0D

>
>// They forgot a const in the prototype of const_timewait...
>
>#define ACE_LACKS_CONST_TIMESPEC_PTR
>

>=0D
>
>#define ACE_LACKS_SYSCALL
>
>#define ACE_HAS_POSIX_TIME
>
>=0D

>
>// Compiler doesn't support static data member templates.
>
>#define ACE_LACKS_STATIC_DATA_MEMBER_TEMPLATES
>

>=0D

>
>// Platform supports System V IPC (most versions of UNIX, but not Win32)
>
>#define ACE_HAS_SYSV_IPC
>

>=0D

>
>// Platform supports recvmsg and sendmsg.
>
>#define ACE_HAS_MSG
>

>=0D

>
>// Compiler/platform contains the <sys/syscall.h> file.
>
>#define ACE_HAS_SYSCALL_H
>

>=0D

>
>// Fixes a problem with HP/UX not wrapping the mmap(2) header files
>
>// with extern "C".
>
>#define ACE_HAS_BROKEN_MMAP_H
>

>=0D

>
>// Prototypes for both signal() and struct sigaction are consistent.
>
>#define ACE_HAS_CONSISTENT_SIGNAL_PROTOTYPES
>

>=0D

>
>// Compiler/platform has correctly prototyped header files.
>
>#define ACE_HAS_CPLUSPLUS_HEADERS
>

>=0D

>
>// Header files lack t_errno for ACE_TLI.
>
>#define ACE_LACKS_T_ERRNO
>

>=0D

>
>// Compiler/platform supports poll().
>
>#define ACE_HAS_POLL
>

>=0D

>
>// Platform supports POSIX O_NONBLOCK semantics.
>
>#define ACE_HAS_POSIX_NONBLOCK
>

>=0D

>
>// Compiler/platform defines the sig_atomic_t typedef
>
>#define ACE_HAS_SIG_ATOMIC_T
>

>=0D

>
>//#define ACE_HAS_THREADS
>
>//#define ACE_HAS_PTHREADS
>
>//#define ACE_MT_SAFE
>
>#define ACE_HAS_SIGINFO_T
>
>//#define ACE_LACKS_PTHREAD_THR_SIGSETMASK
>
>#define ACE_HAS_SETKIND_NP
>
>#define ACE_LACKS_CONDATTR_PSHARED
>
>#define ACE_LACKS_SI_ADDR
>
>#define ACE_LACKS_SETDETACH /* new */
>
>#define ACE_LACKS_KEYDELETE /* new */
>
>//#define ACE_LACKS_THREAD_PROCESS_SCOPING
>
>//#define ACE_LACKS_THREAD_STACK_ADDR
>

>=0D

>
>// Compiler/platform has thread-specific storage
>
>//#define ACE_HAS_THREAD_SPECIFIC_STORAGE
>

>=0D

>
>// Compiler/platform supports struct strbuf.
>
>#define ACE_HAS_STRBUF_T
>

>=0D

>
>// Compiler supports the ssize_t typedef.
>
>#define ACE_HAS_SSIZE_T
>

>=0D

>
>// Defines the page size of the system.
>
>#define ACE_PAGE_SIZE 4096
>

>=0D

>
>// Compiler/platform supports strerror ().
>
>#define ACE_HAS_STRERROR
>

>=0D

>
>// SunOS 4 style prototype.
>
>#define ACE_HAS_SUNOS4_GETTIMEOFDAY
>

>=0D

>
>// HP/UX has an undefined syscall for GETRUSAGE...
>
>#define ACE_HAS_SYSCALL_GETRUSAGE
>

>=0D

>
>// Note, this only works if the flag is set above!
>
>#define ACE_HAS_GETRUSAGE
>

>=0D

>
>// Platform uses int for select() rather than fd_set.
>
>//#define ACE_SELECT_USES_INT
>

>=0D

>
>// Platform has the XLI version of ACE_TLI.
>
>// #define ACE_HAS_XLI
>

>=0D

>
>// Platform supports ACE_TLI tiuser header.
>
>#define ACE_HAS_TIUSER_H
>

>=0D

>
>// Platform provides ACE_TLI function prototypes.
>
>//#define ACE_HAS_TLI_PROTOTYPES
>

>=0D

>
>// Platform supports ACE_TLI.
>
>//#define ACE_HAS_TLI
>

>=0D

>
>// Turns on the tracing feature.
>
>//#if !defined (ACE_NTRACE)
>
>//#define ACE_NTRACE 0 // set to 1 to turn off
>
>//#endif /* ACE_NTRACE */
>

>=0D
>
>=0D
>
>#endif /* ACE_CONFIG_H */
>
>=0D

>
> THE $ACE_ROOT/include/makeinclude/platform_macros.GNU FILE [if you
>
> use a link to a platform-specific file, simply state which one
>
> (unless this isn't used in this case, e.g., with Microsoft Visual
>

> C++)]:=0D
>
> Not available=0D
>
>=0D

>
> CONTENTS OF $ACE_ROOT/bin/MakeProjectCreator/config/default.features
>
> (used by MPC when you generate your own makefiles):
>
> Not available
>

>=0D

>
> AREA/CLASS/EXAMPLE AFFECTED:
>
> [What example failed? What module failed to compile?]
>

> =0D
>
>ACE_Reactor *pReactor
>
>while (1) {
>
> if( pReactor->handle_events()=3D=3D-1)

>
> {
>
> syslog(LOG_ERR, "Error occurred in Reactor with
>errno %d , trying to recover..\n",errno);
>
> }
>

>=0D
>
> }
>
>=0D

>
> Description: The handle_events() function in the above piece of code

>is returning -1.=0D
>
>=0D

>
> DOES THE PROBLEM AFFECT:
>
> COMPILATION?
>
> LINKING?
>
> On Unix systems, did you run make realclean first?
>

> EXECUTION?=0D

>
> OTHER (please specify)?
>
>[Please indicate whether ACE, your application, or both are affected.]
>
>As ACE_Reactor ::handle_events() function returns -1 both ACE and our
>application are affected.
>

>=0D

>
> SYNOPSIS:
>
>[Brief description of the problem]
>
>Our application is a Element management system which is monitoring for
>events and sending requests to a Network Element. The applications
>functions fine for some time around 1 or 2 days and then at some point
>in time, the ACE_Reactor->handle_events() function starts returning -1
>(error scenario) and it does not recover until the whole application is
>killed and restarted. We have tried to print the system error number

>which gets displayed as "9".=0D
>
>=0D

>
> DESCRIPTION:
>
>[Detailed description of problem. Don't just say "<blah>
>
>doesn't work, here's a fix," explain what your program does
>
>to get to the <blah> state. ]
>
>The application is monitoring for events and if there is one particular
>Network element to which the connection is not quite good (i.e. the
>requests made to the Network element from the EMS keep on timing out)
>then this problem is seen to occur. If the connectivity problem is
>rectified, the application doesn't ever get in to the problem state.
>This is the only characterization that was possible. We tried relating
>it to the number of events or Network elements to check if there are any
>resources which the ACE framework is using are getting exhausted, but
>couldn't characterize this.
>

>=0D

>
> REPEAT BY:
>
>[What you did to get the error; include test program or session
>
>transcript if at all possible. ]
>
>If there is a network element to which the connection is not proper (say
>the connecting is lost), then the handle_events() API starts returning
>-1 and the application goes in to a "hung" state because f the while
>loop.
>

>=0D

>
>Since restart of the application solves the problem, we tried restarting
>the Reactor (by using .stop() and .start()) APIs), but this did not
>help. Only once you restart the application, the problem will not be
>seen (for some time again, may be 1 or 2 days).
>

>=0D

>
> SAMPLE FIX/WORKAROUND:
>
>Restart of the application solves the problem but again for some time,
>hence not a solution. The workaround we used was to correct the
>connectivity problem to the Network element from the EMS after which the

>problem was not seen.=0D
>
>=0D
>
>=0D
>
> Regards,
>
> Prashant Poulkar=0D
>
> - - - - - - - - - - - - - - - - - - - - - - - - - - - -=0D

>
> Wipro Technologies, Bangalore
>
> Phone: +91 80 28520408 Extn 3269
>

> - - - - - - - - - - - - - - - - - - - - - - - - - - - -=0D
>
>=0D
>
>=0D

>
>
>The information contained in this electronic message and any attachments
>to this message are intended for the exclusive use of the addressee(s)
>and may contain proprietary, confidential or privileged information. If
>you are not the intended recipient, you should not disseminate,
>distribute or copy this e-mail. Please notify the sender immediately and

>destroy all copies of this message and any attachments.=0D

>
>WARNING: Computer viruses can be transmitted via email. The recipient
>should check this email and any attachments for the presence of viruses.
>The company accepts no liability for any damage caused by any virus
>transmitted by this email.
>
>www.wipro.com

>=0D
>
>
>
>The information contained in this electronic message and any attachments to=
> this message are intended for the exclusive use of the addressee(s) and=
> may contain proprietary, confidential or privileged information. If you=
> are not the intended recipient, you should not disseminate, distribute or=
> copy this e-mail. Please notify the sender immediately and destroy all=
> copies of this message and any attachments.=0D
>
>WARNING: Computer viruses can be transmitted via email. The recipient=
> should check this email and any attachments for the presence of viruses.=
> The company accepts no liability for any damage caused by any virus=
> transmitted by this email.
>=0D
>www.wipro.com
>------_=_NextPart_001_01C6CD04.785ECC6A
>Content-Type: text/html;
> charset="us-ascii"
>Content-Transfer-Encoding: quoted-printable
>
><html xmlns:v=3D"urn:schemas-microsoft-com:vml" xmlns:o=
>=3D"urn:schemas-microsoft-com:office:office" xmlns:w=
>=3D"urn:schemas-microsoft-com:office:word" xmlns:x=
>=3D"urn:schemas-microsoft-com:office:excel" xmlns:st1=
>=3D"urn:schemas-microsoft-com:office:smarttags" xmlns=
>=3D"http://www.w3.org/TR/REC-html40">
>
><head>
><meta http-equiv=3DContent-Type content=3D"text/html; charset=3Dus-ascii">
><meta name=3DGenerator content=3D"Microsoft Word 11 (filtered medium)">
><o:SmartTagType namespaceuri=3D"urn:schemas-microsoft-com:office:smarttags"
> name=3D"place" downloadurl=3D"http://www.5iantlavalamp.com/"/>
><o:SmartTagType namespaceuri=3D"urn:schemas-microsoft-com:office:smarttags"
> name=3D"City" downloadurl=
>=3D"http://www.5iamas-microsoft-com:office:smarttags"/>
><o:SmartTagType namespaceuri=3D"urn:schemas-microsoft-com:office:smarttags"
> name=3D"PersonName" downloadurl=3D"http://www.microsoft.com"/>
>
><style>
>
></style>
>
></head>
>
><body lang=3DEN-US link=3Dblue vlink=3Dpurple>
>
><div class=3DSection1>
>
>=3D'font-size:
>10.0pt;font-family:Arial;color:blue'>Hi,<o:p></o:p>
>
>=3D'font-size:
>10.0pt;font-family:Arial;color:blue'><o:p> </o:p>
>
>=3D'font-size:
>10.0pt;font-family:Arial;color:blue'>We are working for a style=3D'font-weight:bold'>Nortel product which is one of the=
> sponsors
>listed. We are having a very critical issue with the ACE library which we=
> are
>reporting. For one of our Element Management Systems for Telecom networks,=
> we
>are using ACE as the communication framework. Your help will be appreciated=
> in
>addressing this issue. Please let us know, if you need further=
> clarification.<o:p></o:p>
>
>=3D'font-size:10.0pt;
>font-family:Arial'><o:p> </o:p>
>
>=3D'font-size:10.0pt;
>font-family:Arial'>    =3D'font-weight:bold'>ACE
>VERSION: =3D'color:blue'>5.1.17<o:p></o:p>
>
>=3D'font-size:10.0pt;
>font-family:Arial;font-weight:bold'><o:p> </o:p>
>
>=3D'font-size:10.0pt;
>font-family:Arial;font-weight:bold'>    HOST MACHINE and
>OPERATING SYSTEM:<o:p></o:p>
>
>=3D'font-size:
>10.0pt;font-family:Arial;color:blue'>    # uname=
> -a<o:p></o:p>
>
>=3D'font-size:
>10.0pt;font-family:Arial;color:blue'>    HP-UX shivalik=
> B.11.11
>U 9000/785 2007660371 unlimited-user license<o:p></o:p>
>
>=3D'font-size:10.0pt;
>font-family:Arial'><o:p> </o:p>
>
>=3D'font-size:10.0pt;
>font-family:Arial;font-weight:bold'>    TARGET MACHINE and
>OPERATING SYSTEM, if different from HOST:<o:p></o:p>
>
>=3D'font-size:
>10.0pt;font-family:Arial;color:blue'>    # uname=
> -a<o:p></o:p>
>
>=3D'font-size:
>10.0pt;font-family:Arial;color:blue'>    HP-UX shivalik=
> B.11.11
>U 9000/785 2007660371 unlimited-user license<o:p></o:p>
>
>=3D'font-size:10.0pt;
>font-family:Arial'><o:p> </o:p>
>
>=3D'font-size:10.0pt;
>font-family:Arial;font-weight:bold'>    COMPILER NAME AND
>VERSION (AND PATCHLEVEL):<o:p></o:p>
>
>=3D'font-size:10.0pt;
>font-family:Arial'>     style=3D'color:blue'>what /usr/lib/libCsup.sl=
> <o:p></o:p>
>
>=3D'font-size:
>10.0pt;font-family:Arial;color:blue'>    =
> /usr/lib/libCsup.sl:<o:p></o:p>
>
>=3D'font-size:
>10.0pt;font-family:Arial;color:blue'>    
>            &nb=
>sp;     
>HP aC++ B3910B A.03.70 Language Support=
> Library<o:p></o:p>
>
>=3D'font-size:10.0pt;
>font-family:Arial'><o:p> </o:p>
>
>=3D'font-size:10.0pt;
>font-family:Arial;font-weight:bold'>    THE
>$ACE_ROOT/ace/config.h FILE [if you use a link to a=
> platform-<o:p></o:p>
>
>=3D'font-size:10.0pt;
>font-family:Arial;font-weight:bold'>    specific file,=
> simply
>state which one]:<o:p></o:p>
>
>=3D'font-size:
>10.0pt;font-family:Arial;color:blue'>    /* -*- C++ -*-=
> */<o:p></o:p>
>
>=3D'font-size:
>10.0pt;font-family:Arial;color:blue'>// $Id: config-hpux-10.x.h,v 4.5
>1996/12/13 22:30:00 schmidt Exp $<o:p></o:p>
>
>=3D'font-size:
>10.0pt;font-family:Arial;color:blue'><o:p> </o:p>
>
>=3D'font-size:
>10.0pt;font-family:Arial;color:blue'>// The following configuration file is
>designed to work for HP<o:p></o:p>
>
>=3D'font-size:
>10.0pt;font-family:Arial;color:blue'>// platforms running HP/UX 10.x. =
> For
>using HP C++ compiler with<o:p></o:p>
>
>=3D'font-size:
>10.0pt;font-family:Arial;color:blue'>// templates you should have patch
>PHSS_6246.<o:p></o:p>
>
>=3D'font-size:
>10.0pt;font-family:Arial;color:blue'><o:p> </o:p>
>
>=3D'font-size:
>10.0pt;font-family:Arial;color:blue'>#if !defined=
> (ACE_CONFIG_H)<o:p></o:p>
>
>=3D'font-size:
>10.0pt;font-family:Arial;color:blue'>#define=
> ACE_CONFIG_H<o:p></o:p>
>
>=3D'font-size:
>10.0pt;font-family:Arial;color:blue'><o:p> </o:p>
>
>=3D'font-size:
>10.0pt;font-family:Arial;color:blue'>// Compiling for=
> HPUX.<o:p></o:p>
>
>=3D'font-size:
>10.0pt;font-family:Arial;color:blue'>#if !defined=
> (HPUX)<o:p></o:p>
>
>=3D'font-size:
>10.0pt;font-family:Arial;color:blue'>#define=
> HPUX<o:p></o:p>
>
>=3D'font-size:
>10.0pt;font-family:Arial;color:blue'>#endif /* HPUX=
> */<o:p></o:p>
>
>=3D'font-size:
>10.0pt;font-family:Arial;color:blue'><o:p> </o:p>
>
>=3D'font-size:
>10.0pt;font-family:Arial;color:blue'>#define=
> _HPUX_SOURCE<o:p></o:p>
>
>=3D'font-size:
>10.0pt;font-family:Arial;color:blue'>#include /**/=
> <sys/stdsyms.h><o:p></o:p>
>
>=3D'font-size:
>10.0pt;font-family:Arial;color:blue'>#include /**/
><sched.h>          =
>            &nb=
>sp;      
>/*  pthread.h doesn't include this */<o:p></o:p>
>
>=3D'font-size:
>10.0pt;font-family:Arial;color:blue'><o:p> </o:p>
>
>=3D'font-size:
>10.0pt;font-family:Arial;color:blue'>#define=
> ACE_HAS_BROKEN_CONVERSIONS<o:p></o:p>
>
>=3D'font-size:
>10.0pt;font-family:Arial;color:blue'><o:p> </o:p>
>
>=3D'font-size:
>10.0pt;font-family:Arial;color:blue'>// They forgot a const in the=
> prototype of
>const_timewait...<o:p></o:p>
>
>=3D'font-size:
>10.0pt;font-family:Arial;color:blue'>#define=
> ACE_LACKS_CONST_TIMESPEC_PTR<o:p></o:p>
>
>=3D'font-size:
>10.0pt;font-family:Arial;color:blue'><o:p> </o:p>
>
>=3D'font-size:
>10.0pt;font-family:Arial;color:blue'>#define=
> ACE_LACKS_SYSCALL<o:p></o:p>
>
>=3D'font-size:
>10.0pt;font-family:Arial;color:blue'>#define=
> ACE_HAS_POSIX_TIME<o:p></o:p>
>
>=3D'font-size:
>10.0pt;font-family:Arial;color:blue'><o:p> </o:p>
>
>=3D'font-size:
>10.0pt;font-family:Arial;color:blue'>// Compiler doesn't support static=
> data
>member templates.<o:p></o:p>
>
>=3D'font-size:
>10.0pt;font-family:Arial;color:blue'>#define=
> ACE_LACKS_STATIC_DATA_MEMBER_TEMPLATES<o:p></o:p>
>
>=3D'font-size:
>10.0pt;font-family:Arial;color:blue'><o:p> </o:p>
>
>=3D'font-size:
>10.0pt;font-family:Arial;color:blue'>// Platform supports System V IPC=
> (most
>versions of UNIX, but not Win32)<o:p></o:p>
>
>=3D'font-size:
>10.0pt;font-family:Arial;color:blue'>#define=
> ACE_HAS_SYSV_IPC<o:p></o:p>
>
>=3D'font-size:
>10.0pt;font-family:Arial;color:blue'><o:p> </o:p>
>
>=3D'font-size:
>10.0pt;font-family:Arial;color:blue'>// Platform supports recvmsg and=
> sendmsg.<o:p></o:p>
>
>=3D'font-size:
>10.0pt;font-family:Arial;color:blue'>#define=
> ACE_HAS_MSG<o:p></o:p>
>
>=3D'font-size:
>10.0pt;font-family:Arial;color:blue'><o:p> </o:p>
>
>=3D'font-size:
>10.0pt;font-family:Arial;color:blue'>// Compiler/platform contains the
><sys/syscall.h> file.<o:p></o:p>
>
>=3D'font-size:
>10.0pt;font-family:Arial;color:blue'>#define=
> ACE_HAS_SYSCALL_H<o:p></o:p>
>
>=3D'font-size:
>10.0pt;font-family:Arial;color:blue'><o:p> </o:p>
>
>=3D'font-size:
>10.0pt;font-family:Arial;color:blue'>// Fixes a problem with HP/UX not=
> wrapping
>the mmap(2) header files<o:p></o:p>
>
>=3D'font-size:
>10.0pt;font-family:Arial;color:blue'>// with extern=
> "C".<o:p></o:p>
>
>=3D'font-size:
>10.0pt;font-family:Arial;color:blue'>#define=
> ACE_HAS_BROKEN_MMAP_H<o:p></o:p>
>
>=3D'font-size:
>10.0pt;font-family:Arial;color:blue'><o:p> </o:p>
>
>=3D'font-size:
>10.0pt;font-family:Arial;color:blue'>// Prototypes for both signal() and=
> struct
>sigaction are consistent.<o:p></o:p>
>
>=3D'font-size:
>10.0pt;font-family:Arial;color:blue'>#define=
> ACE_HAS_CONSISTENT_SIGNAL_PROTOTYPES<o:p></o:p>
>
>=3D'font-size:
>10.0pt;font-family:Arial;color:blue'><o:p> </o:p>
>
>=3D'font-size:
>10.0pt;font-family:Arial;color:blue'>// Compiler/platform has correctly
>prototyped header files.<o:p></o:p>
>
>=3D'font-size:
>10.0pt;font-family:Arial;color:blue'>#define=
> ACE_HAS_CPLUSPLUS_HEADERS<o:p></o:p>
>
>=3D'font-size:
>10.0pt;font-family:Arial;color:blue'><o:p> </o:p>
>
>=3D'font-size:
>10.0pt;font-family:Arial;color:blue'>// Header files lack t_errno for=
> ACE_TLI.<o:p></o:p>
>
>=3D'font-size:
>10.0pt;font-family:Arial;color:blue'>#define=
> ACE_LACKS_T_ERRNO<o:p></o:p>
>
>=3D'font-size:
>10.0pt;font-family:Arial;color:blue'><o:p> </o:p>
>
>=3D'font-size:
>10.0pt;font-family:Arial;color:blue'>// Compiler/platform supports=
> poll().<o:p></o:p>
>
>=3D'font-size:
>10.0pt;font-family:Arial;color:blue'>#define=
> ACE_HAS_POLL<o:p></o:p>
>
>=3D'font-size:
>10.0pt;font-family:Arial;color:blue'><o:p> </o:p>
>
>=3D'font-size:
>10.0pt;font-family:Arial;color:blue'>// Platform supports POSIX O_NONBLOCK
>semantics.<o:p></o:p>
>
>=3D'font-size:
>10.0pt;font-family:Arial;color:blue'>#define=
> ACE_HAS_POSIX_NONBLOCK<o:p></o:p>
>
>=3D'font-size:
>10.0pt;font-family:Arial;color:blue'><o:p> </o:p>
>
>=3D'font-size:
>10.0pt;font-family:Arial;color:blue'>// Compiler/platform defines the
>sig_atomic_t typedef<o:p></o:p>
>
>=3D'font-size:
>10.0pt;font-family:Arial;color:blue'>#define=
> ACE_HAS_SIG_ATOMIC_T<o:p></o:p>
>
>=3D'font-size:
>10.0pt;font-family:Arial;color:blue'><o:p> </o:p>
>
>=3D'font-size:
>10.0pt;font-family:Arial;color:blue'>//#define=
> ACE_HAS_THREADS<o:p></o:p>
>
>=3D'font-size:
>10.0pt;font-family:Arial;color:blue'>//#define=
> ACE_HAS_PTHREADS<o:p></o:p>
>
>=3D'font-size:
>10.0pt;font-family:Arial;color:blue'>//#define=
> ACE_MT_SAFE<o:p></o:p>
>
>=3D'font-size:
>10.0pt;font-family:Arial;color:blue'>#define=
> ACE_HAS_SIGINFO_T<o:p></o:p>
>
>=3D'font-size:
>10.0pt;font-family:Arial;color:blue'>//#define=
> ACE_LACKS_PTHREAD_THR_SIGSETMASK<o:p></o:p>
>
>=3D'font-size:
>10.0pt;font-family:Arial;color:blue'>#define=
> ACE_HAS_SETKIND_NP<o:p></o:p>
>
>=3D'font-size:
>10.0pt;font-family:Arial;color:blue'>#define=
> ACE_LACKS_CONDATTR_PSHARED<o:p></o:p>
>
>=3D'font-size:
>10.0pt;font-family:Arial;color:blue'>#define=
> ACE_LACKS_SI_ADDR<o:p></o:p>
>
>=3D'font-size:
>10.0pt;font-family:Arial;color:blue'>#define
>ACE_LACKS_SETDETACH         &n=
>bsp;          
>/*  new */<o:p></o:p>
>
>=3D'font-size:
>10.0pt;font-family:Arial;color:blue'>#define
>ACE_LACKS_KEYDELETE         &n=
>bsp;          
>/*  new */<o:p></o:p>
>
>=3D'font-size:
>10.0pt;font-family:Arial;color:blue'>//#define=
> ACE_LACKS_THREAD_PROCESS_SCOPING<o:p></o:p>
>
>=3D'font-size:
>10.0pt;font-family:Arial;color:blue'>//#define=
> ACE_LACKS_THREAD_STACK_ADDR<o:p></o:p>
>
>=3D'font-size:
>10.0pt;font-family:Arial;color:blue'><o:p> </o:p>
>
>=3D'font-size:
>10.0pt;font-family:Arial;color:blue'>// Compiler/platform has=
> thread-specific
>storage<o:p></o:p>
>
>=3D'font-size:
>10.0pt;font-family:Arial;color:blue'>//#define=
> ACE_HAS_THREAD_SPECIFIC_STORAGE<o:p></o:p>
>
>=3D'font-size:
>10.0pt;font-family:Arial;color:blue'><o:p> </o:p>
>
>=3D'font-size:
>10.0pt;font-family:Arial;color:blue'>// Compiler/platform supports struct
>strbuf.<o:p></o:p>
>
>=3D'font-size:
>10.0pt;font-family:Arial;color:blue'>#define=
> ACE_HAS_STRBUF_T<o:p></o:p>
>
>=3D'font-size:
>10.0pt;font-family:Arial;color:blue'><o:p> </o:p>
>
>=3D'font-size:
>10.0pt;font-family:Arial;color:blue'>// Compiler supports the ssize_t=
> typedef.<o:p></o:p>
>
>=3D'font-size:
>10.0pt;font-family:Arial;color:blue'>#define=
> ACE_HAS_SSIZE_T<o:p></o:p>
>
>=3D'font-size:
>10.0pt;font-family:Arial;color:blue'><o:p> </o:p>
>
>=3D'font-size:
>10.0pt;font-family:Arial;color:blue'>// Defines the page size of the=
> system.<o:p></o:p>
>
>=3D'font-size:
>10.0pt;font-family:Arial;color:blue'>#define ACE_PAGE_SIZE=
> 4096<o:p></o:p>
>
>=3D'font-size:
>10.0pt;font-family:Arial;color:blue'><o:p> </o:p>
>
>=3D'font-size:
>10.0pt;font-family:Arial;color:blue'>// Compiler/platform supports strerror=
> ().<o:p></o:p>
>
>=3D'font-size:
>10.0pt;font-family:Arial;color:blue'>#define=
> ACE_HAS_STRERROR<o:p></o:p>
>
>=3D'font-size:
>10.0pt;font-family:Arial;color:blue'><o:p> </o:p>
>
>=3D'font-size:
>10.0pt;font-family:Arial;color:blue'>// SunOS 4 style=
> prototype.<o:p></o:p>
>
>=3D'font-size:
>10.0pt;font-family:Arial;color:blue'>#define=
> ACE_HAS_SUNOS4_GETTIMEOFDAY<o:p></o:p>
>
>=3D'font-size:
>10.0pt;font-family:Arial;color:blue'><o:p> </o:p>
>
>=3D'font-size:
>10.0pt;font-family:Arial;color:blue'>// HP/UX has an undefined syscall for
>GETRUSAGE...<o:p></o:p>
>
>=3D'font-size:
>10.0pt;font-family:Arial;color:blue'>#define=
> ACE_HAS_SYSCALL_GETRUSAGE<o:p></o:p>
>
>=3D'font-size:
>10.0pt;font-family:Arial;color:blue'><o:p> </o:p>
>
>=3D'font-size:
>10.0pt;font-family:Arial;color:blue'>// Note, this only works if the flag=
> is
>set above!<o:p></o:p>
>
>=3D'font-size:
>10.0pt;font-family:Arial;color:blue'>#define=
> ACE_HAS_GETRUSAGE<o:p></o:p>
>
>=3D'font-size:
>10.0pt;font-family:Arial;color:blue'><o:p> </o:p>
>
>=3D'font-size:
>10.0pt;font-family:Arial;color:blue'>// Platform uses int for select()=
> rather
>than fd_set.<o:p></o:p>
>
>=3D'font-size:
>10.0pt;font-family:Arial;color:blue'>//#define=
> ACE_SELECT_USES_INT<o:p></o:p>
>
>=3D'font-size:
>10.0pt;font-family:Arial;color:blue'><o:p> </o:p>
>
>=3D'font-size:
>10.0pt;font-family:Arial;color:blue'>// Platform has the XLI version of
>ACE_TLI.<o:p></o:p>
>
>=3D'font-size:
>10.0pt;font-family:Arial;color:blue'>// #define=
> ACE_HAS_XLI<o:p></o:p>
>
>=3D'font-size:
>10.0pt;font-family:Arial;color:blue'><o:p> </o:p>
>
>=3D'font-size:
>10.0pt;font-family:Arial;color:blue'>// Platform supports ACE_TLI tiuser
>header.<o:p></o:p>
>
>=3D'font-size:
>10.0pt;font-family:Arial;color:blue'>#define=
> ACE_HAS_TIUSER_H<o:p></o:p>
>
>=3D'font-size:
>10.0pt;font-family:Arial;color:blue'><o:p> </o:p>
>
>=3D'font-size:
>10.0pt;font-family:Arial;color:blue'>// Platform provides ACE_TLI function
>prototypes.<o:p></o:p>
>
>=3D'font-size:
>10.0pt;font-family:Arial;color:blue'>//#define=
> ACE_HAS_TLI_PROTOTYPES<o:p></o:p>
>
>=3D'font-size:
>10.0pt;font-family:Arial;color:blue'><o:p> </o:p>
>
>=3D'font-size:
>10.0pt;font-family:Arial;color:blue'>// Platform supports=
> ACE_TLI.<o:p></o:p>
>
>=3D'font-size:
>10.0pt;font-family:Arial;color:blue'>//#define=
> ACE_HAS_TLI<o:p></o:p>
>
>=3D'font-size:
>10.0pt;font-family:Arial;color:blue'><o:p> </o:p>
>
>=3D'font-size:
>10.0pt;font-family:Arial;color:blue'>// Turns on the tracing=
> feature.<o:p></o:p>
>
>=3D'font-size:
>10.0pt;font-family:Arial;color:blue'>//#if !defined=
> (ACE_NTRACE)<o:p></o:p>
>
>=3D'font-size:
>10.0pt;font-family:Arial;color:blue'>//#define ACE_NTRACE 0   //=
> set
>to 1 to turn off<o:p></o:p>
>
>=3D'font-size:
>10.0pt;font-family:Arial;color:blue'>//#endif /* ACE_NTRACE=
> */<o:p></o:p>
>
>=3D'font-size:
>10.0pt;font-family:Arial;color:blue'><o:p> </o:p>
>
>=3D'font-size:
>10.0pt;font-family:Arial;color:blue'><o:p> </o:p>
>
>=3D'font-size:
>10.0pt;font-family:Arial;color:blue'>#endif /* ACE_CONFIG_H=
> */<o:p></o:p>
>
>=3D'font-size:10.0pt;
>font-family:Arial'><o:p> </o:p>
>
>=3D'font-size:10.0pt;
>font-family:Arial;font-weight:bold'>    THE
>$ACE_ROOT/include/makeinclude/platform_macros.GNU FILE [if=
> you<o:p></o:p>
>
>=3D'font-size:10.0pt;
>font-family:Arial;font-weight:bold'>    use a link to a
>platform-specific file, simply state which=
> one<o:p></o:p>
>
>=3D'font-size:10.0pt;
>font-family:Arial;font-weight:bold'>    (unless this isn't=
> used
>in this case, e.g., with Microsoft Visual<o:p></o:p>
>
>=3D'font-size:10.0pt;
>font-family:Arial;font-weight:bold'>    C++)]:=
> <o:p></o:p>
>
>=3D'font-size:
>10.0pt;font-family:Arial;color:blue'>     Not available=
> <o:p></o:p>
>
>=3D'font-size:10.0pt;
>font-family:Arial'><o:p> </o:p>
>
>=3D'font-size:10.0pt;
>font-family:Arial;font-weight:bold'>    CONTENTS OF
>$ACE_ROOT/bin/MakeProjectCreator/config/default.features<o:p></o:p><=
>/font>
>
>=3D'font-size:10.0pt;
>font-family:Arial;font-weight:bold'>    (used by MPC when=
> you
>generate your own makefiles):<o:p></o:p>
>
>=3D'font-size:
>10.0pt;font-family:Arial;color:blue'>    Not=
> available<o:p></o:p>
>
>=3D'font-size:10.0pt;
>font-family:Arial'><o:p> </o:p>
>
>=3D'font-size:10.0pt;
>font-family:Arial;font-weight:bold'>    AREA/CLASS/EXAMPLE
>AFFECTED:<o:p></o:p>
>
>=3D'font-size:10.0pt;
>font-family:Arial;font-weight:bold'>    [What example
>failed?  What module failed to=
> compile?]<o:p></o:p>
>
>=3D'font-size:10.0pt;
>font-family:Arial;font-weight:bold'>   =
> <o:p></o:p>
>
> face=3DArial>style=3D'font-size:10.0pt;font-family:Arial;color:blue'>ACE_Reactor=
> *pReactor<o:p></o:p>
>
> face=3DArial>style=3D'font-size:10.0pt;font-family:Arial;color:blue'>while (1)=
> {<o:p></o:p>
>
>=3D'font-size:
>10.0pt;font-family:Arial;color:blue'> 
>            if(
>pReactor->handle_events()=3D=3D-1)<o:p></o:p>
>
>=3D'font-size:
>10.0pt;font-family:Arial;color:blue'> 
>         =
>   {<o:p></o:p>
>
>=3D'font-size:
>10.0pt;font-family:Arial;color:blue'>      &n=
>bsp;
>            &nb=
>sp;  
>syslog(LOG_ERR, "Error occurred in Reactor  with errno %d ,=
> trying to
>recover..\n",errno);<o:p></o:p>
>
>=3D'font-size:
>10.0pt;font-family:Arial;color:blue'>   
>           }<o:p></o=
>:p>
>
>=3D'font-size:
>10.0pt;font-family:Arial;color:blue'><o:p> </o:p>
>
>=3D'font-size:
>10.0pt;font-family:Arial;color:blue'> 
>         =
> }<o:p></o:p>
>
>=3D'font-size:
>10.0pt;font-family:Arial;color:blue'><o:p> </o:p>
>
>=3D'font-size:
>10.0pt;font-family:Arial;color:blue'>    Description: The
>handle_events() function in the above piece of code is returning -1.=
> <o:p></o:p>
>
>=3D'font-size:10.0pt;
>font-family:Arial'><o:p> </o:p>
>
>=3D'font-size:10.0pt;
>font-family:Arial;font-weight:bold'>    DOES THE PROBLEM=
> AFFECT:<o:p></o:p>
>
>=3D'font-size:10.0pt;
>font-family:Arial;font-weight:bold'>      &nb=
>sp;
>COMPILATION?<o:p></o:p>
>
>=3D'font-size:10.0pt;
>font-family:Arial;font-weight:bold'>      &nb=
>sp;
>LINKING?<o:p></o:p>
>
>=3D'font-size:10.0pt;
>font-family:Arial;font-weight:bold'>      &nb=
>sp;    
>On Unix systems, did you run make realclean=
> first?<o:p></o:p>
>
>=3D'font-size:10.0pt;
>font-family:Arial;font-weight:bold'>      &nb=
>sp;
>EXECUTION? <o:p></o:p>
>
>=3D'font-size:10.0pt;
>font-family:Arial;font-weight:bold'>      &nb=
>sp;
>OTHER (please specify)?<o:p></o:p>
>
>=3D'font-size:10.0pt;
>font-family:Arial;font-weight:bold'>[Please indicate whether ACE, your
>application, or both are affected.]<o:p></o:p>
>
>=3D'font-size:
>10.0pt;font-family:Arial;color:blue'>As ACE_Reactor ::handle_events()=
> function
>returns -1 both ACE and our application are=
> affected.<o:p></o:p>
>
>=3D'font-size:10.0pt;
>font-family:Arial'><o:p> </o:p>
>
>=3D'font-size:10.0pt;
>font-family:Arial;font-weight:bold'>   =
> SYNOPSIS:<o:p></o:p>
>
>=3D'font-size:10.0pt;
>font-family:Arial;font-weight:bold'>[Brief description of the=
> problem]<o:p></o:p>
>
>=3D'font-size:
>10.0pt;font-family:Arial;color:blue'>Our application is a Element=

> management
>system which is monitoring for events and sending requests to a Network

>Element. The applications functions fine for some time around 1 or 2 days=
> and
>then at some point in time, the ACE_Reactor->handle_events() function=

> starts
>returning -1 (error scenario) and it does not recover until the whole

>application is killed and restarted. We have tried to print the system=
> error
>number which gets displayed as “9”.=
> <o:p></o:p>
>
>=3D'font-size:10.0pt;
>font-family:Arial'><o:p> </o:p>
>
>=3D'font-size:10.0pt;
>font-family:Arial;font-weight:bold'>   =
> DESCRIPTION:<o:p></o:p>
>
>=3D'font-size:10.0pt;
>font-family:Arial;font-weight:bold'>[Detailed description of problem. 
>Don't just say "<blah><o:p></o:p>
>
>=3D'font-size:10.0pt;
>font-family:Arial;font-weight:bold'>doesn't work, here's a fix,"=
> explain
>what your program does<o:p></o:p>
>
>=3D'font-size:10.0pt;
>font-family:Arial;font-weight:bold'>to get to the <blah> state.=
> ]<o:p></o:p>
>
>=3D'font-size:
>10.0pt;font-family:Arial;color:blue'>The application is monitoring for=
> events
>and if there is one particular Network element to which the connection is=
> not
>quite good (i.e. the requests made to the Network element from the=
> <st1:place
>w:st=3D"on">EMS</st1:place> keep on timing out) then this problem is seen=
> to
>occur. If the connectivity problem is rectified, the application=
> doesn’t
>ever get in to the problem state. This is the only characterization that=
> was
>possible. We tried relating it to the number of events or Network elements=
> to
>check if there are any resources which the ACE framework is using are=
> getting
>exhausted, but couldn’t characterize=
> this.<o:p></o:p>
>
>=3D'font-size:10.0pt;
>font-family:Arial;font-weight:bold'><o:p> </o:p>
>
>=3D'font-size:10.0pt;
>font-family:Arial;font-weight:bold'>    REPEAT=
> BY:<o:p></o:p>
>
>=3D'font-size:
>10.0pt;font-family:Arial;color:blue'>[What you did to get the error;=
> include
>test program or session<o:p></o:p>
>
>=3D'font-size:
>10.0pt;font-family:Arial;color:blue'>transcript if at all possible.=
>  ]<o:p></o:p>
>
>=3D'font-size:
>10.0pt;font-family:Arial;color:blue'>If there is a network element to which=
> the
>connection is not proper (say the connecting is lost), then the=
> handle_events()
>API starts returning -1 and the application goes in to a “hung”
>state because f the while loop.<o:p></o:p>
>
>=3D'font-size:
>10.0pt;font-family:Arial;color:blue'><o:p> </o:p>
>
>=3D'font-size:
>10.0pt;font-family:Arial;color:blue'>Since restart of the application=
> solves
>the problem, we tried restarting the Reactor (by using .stop() and=

> .start())
>APIs), but this did not help. Only once you restart the application, the

>problem will not be seen (for some time again, may be 1 or 2=
> days).<o:p></o:p>
>
>=3D'font-size:10.0pt;
>font-family:Arial;font-weight:bold'><o:p> </o:p>
>
>=3D'font-size:10.0pt;
>font-family:Arial;font-weight:bold'>    SAMPLE=
> FIX/WORKAROUND:<o:p></o:p>
>
>=3D'font-size:
>10.0pt;font-family:Arial;color:blue'>Restart of the application solves the
>problem but again for some time, hence not a solution. The workaround we=
> used
>was to correct the connectivity problem to the Network element from the=
> <st1:place
>w:st=3D"on">EMS</st1:place> after which the problem was not seen.=
> <o:p></o:p>
>
>=3D'font-size:10.0pt;
>font-family:Arial'><o:p> </o:p>
>
>=3D'font-size:
>12.0pt'> <o:p></o:p>
>
>=3D'font-size:
>12.0pt'>  Italic">style=3D'font-family:"Harlow Solid=
> Italic";color:blue'>Regards,<o:p></o:p>
>
> Italic">style=3D'font-size:12.0pt;font-family:"Harlow Solid=
> Italic";color:#0080FF'> <st1:PersonName
>w:st=3D"on">=3D'font-family:
> Magneto;color:#8080FF;font-weight:bold'>Prashant=
> Poulkar</st1:PersonName>color=3Dblue face=3DMagneto>=3D'font-family:Magneto;color:blue;font-weight:
>bold'> <o:p></o:p>
>
>=3DGeorgia>style=
>=3D'font-size:10.0pt;font-family:Georgia;color:#8080FF'> </font=
>>size=3D2 color=3D"#8080ff" face=3D"MS Sans Serif">=3D'font-size:10.0pt;
>font-family:"MS Sans Serif";color:#8080FF'>- - - - - - - - - - - - - - - -=
> - -
>- - - - - - - - - - <o:p></o:p>
>
>style=3D'font-size:9.0pt;font-family:Arial;color:#8080FF'> Wipro=
> Technologies,
><st1:place w:st=3D"on"><st1:City w:st=
>=3D"on">Bangalore</st1:City></st1:place>size=3D1><o:p></o:p>
>
>style=3D'font-size:9.0pt;font-family:Arial;color:#8080FF'> Phone: +91=
> 80
>28520408 Extn 3269=3D'font-size:9.0pt'><o:p></o:p>
>
>=3DGeorgia>style=
>=3D'font-size:10.0pt;font-family:Georgia;color:#8080FF'> </font=
>>size=3D2 color=3D"#8080ff" face=3D"MS Sans Serif">=3D'font-size:10.0pt;
>font-family:"MS Sans Serif";color:#8080FF'>- - - - - - - - - - - - - - - -=
> - -
>- - - - - - - - - - <o:p></o:p>
>
>=3D'font-size:
>12.0pt'> <o:p></o:p>
>
>=3D'font-size:
>12.0pt'><o:p> </o:p>
>
></div>
>
></body>
>
></html>
><table><tr><td bgcolor=3D#ffffff> 
>The information contained in this electronic message and any attachments to=
> this message are intended for the exclusive use of the addressee(s) and=
> may contain proprietary, confidential or privileged information. If you=
> are not the intended recipient, you should not disseminate, distribute or=
> copy this e-mail. Please notify the sender immediately and destroy all=
> copies of this message and any attachments. 
> 
>WARNING: Computer viruses can be transmitted via email. The recipient=
> should check this email and any attachments for the presence of viruses.=
> The company accepts no liability for any damage caused by any virus=
> transmitted by this email. 
> 
>www.wipro.com 
></td></tr></table>
><table><tr><td bgcolor=3D#ffffff> 
>The information contained in this electronic message and any attachments to=
> this message are intended for the exclusive use of the addressee(s) and=
> may contain proprietary, confidential or privileged information. If you=
> are not the intended recipient, you should not disseminate, distribute or=
> copy this e-mail. Please notify the sender immediately and destroy all=
> copies of this message and any attachments. 
> 
>WARNING: Computer viruses can be transmitted via email. The recipient=
> should check this email and any attachments for the presence of viruses.=
> The company accepts no liability for any damage caused by any virus=
> transmitted by this email. 
> 
>www.wipro.com 
></td></tr></table>
>------_=_NextPart_001_01C6CD04.785ECC6A--
>

--
Dr. Douglas C. Schmidt Professor and Associate Chair
Electrical Engineering and Computer Science TEL: (615) 343-8197
Vanderbilt University WEB: www.dre.vanderbilt.edu/~schmidt
Nashville, TN 37203 NET: d.sc...@vanderbilt.edu

prashant...@wipro.com

unread,

Aug 31, 2006, 10:18:24 AM8/31/06

to sch...@dre.vanderbilt.edu, ace-...@cse.wustl.edu

Hi Douglas,

Thanks a lot for the information.
Any specific reason why ACE+TAO version 5.2 and why not version 5.5?

Regards,
Prashant Poulkar

- - - - - - - - - - - - - - - - - - - - - - - - - - - -

Wipro Technologies, Bangalore
Phone: +91 80 28520408 Extn 3269

-----Original Message-----
From: Douglas C. Schmidt [mailto:sch...@dre.vanderbilt.edu]
Sent: Thursday, August 31, 2006 7:33 PM
To: Prashant Poulkar (WT01 - Broadband Networks); ace-...@cse.wustl.edu
Subject: Re: [ace-bugs] ACE_Reactor: handle_events API returns -1

Douglas C. Schmidt

unread,

Aug 31, 2006, 10:20:09 AM8/31/06

to prashant...@wipro.com, ace-...@cse.wustl.edu

Hi Prashant,

> Thanks a lot for the information.
> Any specific reason why ACE+TAO version 5.2 and why not version 5.5?

The "x.5.2" below refers to "ACE 5.5.2, TAO 1.5.2, and CIAO 0.5.2",
which is why we refer to it as "x.5.2". This is the latest beta version
available at

http://deuce.doc.wustl.edu/Download.html

Thanks,

Doug

Steve Huston

unread,

Aug 31, 2006, 11:29:24 AM8/31/06

to prashant...@wipro.com, ace-...@cse.wustl.edu, rames...@wipro.com

It looks like there's a problem recovering from bad file descriptors in that version of ACE. Not surprising, since there was no HP-UX 11 support in ACE 5.1...

I recommend that you upgrade to ACE 5.5. If this is not possible, we can work on fixes for you - please contact me if you'd like to discuss the constraints and issues facing your project and how we may be able to help.

Best regards,

-Steve

--
Steve Huston, Riverace Corporation
Next public ACE training October 17-20, 2006!
See http://www.riverace.com/training.htm

-----Original Message-----
From: ace-bugs...@cse.wustl.edu [mailto:ace-bugs...@cse.wustl.edu] On Behalf Of prashant...@wipro.com
Sent: Thursday, August 31, 2006 9:53 AM
To: ace-...@cse.wustl.edu
Cc: rames...@wipro.com

Subject: Re: [ace-bugs] ACE_Reactor: handle_events API returns -1

Hi,

We are working for a Nortel product which is one of the sponsors listed. We are having a very critical issue with the ACE library which we are reporting. For one of our Element Management Systems for Telecom networks, we are using ACE as the communication framework. Your help will be appreciated in addressing this issue. Please let us know, if you need further clarification.

ACE VERSION: 5.1.17

    HOST MACHINE and OPERATING SYSTEM:

    # uname -a

    HP-UX shivalik B.11.11 U 9000/785 2007660371 unlimited-user license

TARGET MACHINE and OPERATING SYSTEM, if different from HOST:

# uname -a

HP-UX shivalik B.11.11 U 9000/785 2007660371 unlimited-user license

    COMPILER NAME AND VERSION (AND PATCHLEVEL):

    what /usr/lib/libCsup.sl

     /usr/lib/libCsup.sl:

HP aC++ B3910B A.03.70 Language Support Library

THE $ACE_ROOT/ace/config.h FILE [if you use a link to a platform-

specific file, simply state which one]:

/* -*- C++ -*- */

// $Id: config-hpux-10.x.h,v 4.5 1996/12/13 22:30:00 schmidt Exp $

// The following configuration file is designed to work for HP

// platforms running HP/UX 10.x. For using HP C++ compiler with

// templates you should have patch PHSS_6246.

#if !defined (ACE_CONFIG_H)

#define ACE_CONFIG_H

// Compiling for HPUX.

#if !defined (HPUX)

#define HPUX

#endif /* HPUX */

#define _HPUX_SOURCE

#include /**/ <sys/stdsyms.h>

#include /**/ <sched.h> /* pthread.h doesn't include this */

#define ACE_HAS_BROKEN_CONVERSIONS

// They forgot a const in the prototype of const_timewait...

#define ACE_LACKS_CONST_TIMESPEC_PTR

#define ACE_LACKS_SYSCALL

#define ACE_HAS_POSIX_TIME

// Compiler doesn't support static data member templates.

#define ACE_LACKS_STATIC_DATA_MEMBER_TEMPLATES

// Platform supports System V IPC (most versions of UNIX, but not Win32)

#define ACE_HAS_SYSV_IPC

// Platform supports recvmsg and sendmsg.

#define ACE_HAS_MSG

// Compiler/platform contains the <sys/syscall.h> file.

#define ACE_HAS_SYSCALL_H

// Fixes a problem with HP/UX not wrapping the mmap(2) header files

// with extern "C".

#define ACE_HAS_BROKEN_MMAP_H

// Prototypes for both signal() and struct sigaction are consistent.

#define ACE_HAS_CONSISTENT_SIGNAL_PROTOTYPES

// Compiler/platform has correctly prototyped header files.

#define ACE_HAS_CPLUSPLUS_HEADERS

// Header files lack t_errno for ACE_TLI.

#define ACE_LACKS_T_ERRNO

// Compiler/platform supports poll().

#define ACE_HAS_POLL

// Platform supports POSIX O_NONBLOCK semantics.

#define ACE_HAS_POSIX_NONBLOCK

// Compiler/platform defines the sig_atomic_t typedef

#define ACE_HAS_SIG_ATOMIC_T

//#define ACE_HAS_THREADS

//#define ACE_HAS_PTHREADS

//#define ACE_MT_SAFE

#define ACE_HAS_SIGINFO_T

//#define ACE_LACKS_PTHREAD_THR_SIGSETMASK

#define ACE_HAS_SETKIND_NP

#define ACE_LACKS_CONDATTR_PSHARED

#define ACE_LACKS_SI_ADDR

#define ACE_LACKS_SETDETACH /* new */

#define ACE_LACKS_KEYDELETE /* new */

//#define ACE_LACKS_THREAD_PROCESS_SCOPING

//#define ACE_LACKS_THREAD_STACK_ADDR

// Compiler/platform has thread-specific storage

//#define ACE_HAS_THREAD_SPECIFIC_STORAGE

// Compiler/platform supports struct strbuf.

#define ACE_HAS_STRBUF_T

// Compiler supports the ssize_t typedef.

#define ACE_HAS_SSIZE_T

// Defines the page size of the system.

#define ACE_PAGE_SIZE 4096

// Compiler/platform supports strerror ().

#define ACE_HAS_STRERROR

// SunOS 4 style prototype.

#define ACE_HAS_SUNOS4_GETTIMEOFDAY

// HP/UX has an undefined syscall for GETRUSAGE...

#define ACE_HAS_SYSCALL_GETRUSAGE

// Note, this only works if the flag is set above!

#define ACE_HAS_GETRUSAGE

// Platform uses int for select() rather than fd_set.

//#define ACE_SELECT_USES_INT

// Platform has the XLI version of ACE_TLI.

// #define ACE_HAS_XLI

// Platform supports ACE_TLI tiuser header.

#define ACE_HAS_TIUSER_H

// Platform provides ACE_TLI function prototypes.

//#define ACE_HAS_TLI_PROTOTYPES

// Platform supports ACE_TLI.

//#define ACE_HAS_TLI

// Turns on the tracing feature.

//#if !defined (ACE_NTRACE)

//#define ACE_NTRACE 0 // set to 1 to turn off

//#endif /* ACE_NTRACE */

#endif /* ACE_CONFIG_H */

    THE $ACE_ROOT/include/makeinclude/platform_macros.GNU FILE [if you

    use a link to a platform-specific file, simply state which one

    (unless this isn't used in this case, e.g., with Microsoft Visual

    C++)]:

Not available

    CONTENTS OF $ACE_ROOT/bin/MakeProjectCreator/config/default.features

    (used by MPC when you generate your own makefiles):

    Not available

AREA/CLASS/EXAMPLE AFFECTED:

[What example failed? What module failed to compile?]

ACE_Reactor *pReactor

while (1) {

if( pReactor->handle_events()==-1)

            {

                        syslog(LOG_ERR, "Error occurred in Reactor with errno %d , trying to recover..\n",errno);

               }

          }

Description: The handle_events() function in the above piece of code is returning -1.

DOES THE PROBLEM AFFECT:

        COMPILATION?

        LINKING?

            On Unix systems, did you run make realclean first?

        EXECUTION?

OTHER (please specify)?

[Please indicate whether ACE, your application, or both are affected.]

As ACE_Reactor ::handle_events() function returns -1 both ACE and our application are affected.

SYNOPSIS:

[Brief description of the problem]

Our application is a Element management system which is monitoring for events and sending requests to a Network Element. The applications functions fine for some time around 1 or 2 days and then at some point in time, the ACE_Reactor->handle_events() function starts returning -1 (error scenario) and it does not recover until the whole application is killed and restarted. We have tried to print the system error number which gets displayed as “9”.

DESCRIPTION:

[Detailed description of problem. Don't just say "<blah>

doesn't work, here's a fix," explain what your program does

to get to the <blah> state. ]

The application is monitoring for events and if there is one particular Network element to which the connection is not quite good (i.e. the requests made to the Network element from the EMS keep on timing out) then this problem is seen to occur. If the connectivity problem is rectified, the application doesn’t ever get in to the problem state. This is the only characterization that was possible. We tried relating it to the number of events or Network elements to check if there are any resources which the ACE framework is using are getting exhausted, but couldn’t characterize this.

REPEAT BY:

[What you did to get the error; include test program or session

transcript if at all possible. ]

If there is a network element to which the connection is not proper (say the connecting is lost), then the handle_events() API starts returning -1 and the application goes in to a “hung” state because f the while loop.

Since restart of the application solves the problem, we tried restarting the Reactor (by using .stop() and .start()) APIs), but this did not help. Only once you restart the application, the problem will not be seen (for some time again, may be 1 or 2 days).

SAMPLE FIX/WORKAROUND:

Restart of the application solves the problem but again for some time, hence not a solution. The workaround we used was to correct the connectivity problem to the Network element from the EMS after which the problem was not seen.

Regards,

Prashant Poulkar

- - - - - - - - - - - - - - - - - - - - - - - - - - - -

Wipro Technologies, Bangalore

Phone: +91 80 28520408 Extn 3269

- - - - - - - - - - - - - - - - - - - - - - - - - - - -

The information contained in this electronic message and any attachments to this message are intended for the exclusive use of the addressee(s) and may contain proprietary, confidential or privileged information. If you are not the intended recipient, you should not disseminate, distribute or copy this e-mail. Please notify the sender immediately and destroy all copies of this message and any attachments.

WARNING: Computer viruses can be transmitted via email. The recipient should check this email and any attachments for the presence of viruses. The company accepts no liability for any damage caused by any virus transmitted by this email.

www.wipro.com

0 new messages