From andy at codesuccess.com Thu Dec 6 19:48:26 2007 From: andy at codesuccess.com (Andy Grove) Date: Thu, 6 Dec 2007 19:48:26 -0000 Subject: [Rhm-users] Which version of RHM C++ broker is compatible with Apache QPID M2? Message-ID: <028201c83840$fd0959d0$f71c0d70$@com> Hi, I've been using Apache QPID for a while now as part of a solution I'm working on and I would like to be able to start testing the RHM C++ broker instead of the Apache QPID Java broker. Previously when I tried this there was a compatibility issue between the QPID client and the RHM broker. Is there now a version of the RHM C++ broker that is compatible with the recent QPID M2 release? Thanks, Andy. -------------- next part -------------- An HTML attachment was scrubbed... URL: From rattapat at redhat.com Thu Dec 6 20:03:41 2007 From: rattapat at redhat.com (Rajith Attapattu) Date: Thu, 06 Dec 2007 15:03:41 -0500 Subject: [Rhm-users] Which version of RHM C++ broker is compatible with Apache QPID M2? In-Reply-To: <028201c83840$fd0959d0$f71c0d70$@com> References: <028201c83840$fd0959d0$f71c0d70$@com> Message-ID: <1196971421.4135.46.camel@localhost.localdomain> Andy, The M2 release is based on the AMQP 0-8 spec and was done based on the M2 branch. RHM c++ broker and the clients are based on the 0-10 version of the AMQP spec (which will be released shortly). This is based on the Apache Qpid trunk. Along with the RHM c++ broker we have clients in c++, java (jms) and python that will work this broker. Regards, Rajith On Thu, 2007-12-06 at 19:48 +0000, Andy Grove wrote: > Hi, > > > > I?ve been using Apache QPID for a while now as part of a solution I?m > working on and I would like to be able to start testing the RHM C++ > broker instead of the Apache QPID Java broker. Previously when I tried > this there was a compatibility issue between the QPID client and the > RHM broker. Is there now a version of the RHM C++ broker that is > compatible with the recent QPID M2 release? > > > > Thanks, > > > > Andy. > > > _______________________________________________ > Rhm-users mailing list > Rhm-users at redhat.com > https://www.redhat.com/mailman/listinfo/rhm-users From andy at codesuccess.com Thu Dec 6 21:05:59 2007 From: andy at codesuccess.com (Andy Grove) Date: Thu, 6 Dec 2007 21:05:59 -0000 Subject: [Rhm-users] Which version of RHM C++ broker is compatible with Apache QPID M2? In-Reply-To: <1196971421.4135.46.camel@localhost.localdomain> References: <028201c83840$fd0959d0$f71c0d70$@com> <1196971421.4135.46.camel@localhost.localdomain> Message-ID: <02ae01c8384b$ceb28dd0$6c17a970$@com> Thanks for the quick response. What's the quickest way to get up and running with a Java client? Do I need to check the source out of subversion and build it or should I use one of the source RPMs that are available for download? Are there any binary distributions for the client? Thanks, Andy. -----Original Message----- From: Rajith Attapattu [mailto:rattapat at redhat.com] Sent: 06 December 2007 20:04 To: Andy Grove Cc: rhm-users at redhat.com Subject: Re: [Rhm-users] Which version of RHM C++ broker is compatible with Apache QPID M2? Andy, The M2 release is based on the AMQP 0-8 spec and was done based on the M2 branch. RHM c++ broker and the clients are based on the 0-10 version of the AMQP spec (which will be released shortly). This is based on the Apache Qpid trunk. Along with the RHM c++ broker we have clients in c++, java (jms) and python that will work this broker. Regards, Rajith On Thu, 2007-12-06 at 19:48 +0000, Andy Grove wrote: > Hi, > > > > I?ve been using Apache QPID for a while now as part of a solution I?m > working on and I would like to be able to start testing the RHM C++ > broker instead of the Apache QPID Java broker. Previously when I tried > this there was a compatibility issue between the QPID client and the > RHM broker. Is there now a version of the RHM C++ broker that is > compatible with the recent QPID M2 release? > > > > Thanks, > > > > Andy. > > > _______________________________________________ > Rhm-users mailing list > Rhm-users at redhat.com > https://www.redhat.com/mailman/listinfo/rhm-users From cctrieloff at redhat.com Thu Dec 6 21:16:22 2007 From: cctrieloff at redhat.com (Carl Trieloff) Date: Thu, 06 Dec 2007 16:16:22 -0500 Subject: [Rhm-users] Which version of RHM C++ broker is compatible with Apache QPID M2? In-Reply-To: <02ae01c8384b$ceb28dd0$6c17a970$@com> References: <028201c83840$fd0959d0$f71c0d70$@com> <1196971421.4135.46.camel@localhost.localdomain> <02ae01c8384b$ceb28dd0$6c17a970$@com> Message-ID: <475866A6.4040902@redhat.com> Full instructions to do it all from rpm's should sync live a bit later today. A download page will appear on www.redhat.com/mrg We can also post you the yum repo URL once the sync up is complete. Carl. Andy Grove wrote: > Thanks for the quick response. What's the quickest way to get up and running with a Java client? Do I need to check the source out of subversion and build it or should I use one of the source RPMs that are available for download? Are there any binary distributions for the client? > > Thanks, > > Andy. > > -----Original Message----- > From: Rajith Attapattu [mailto:rattapat at redhat.com] > Sent: 06 December 2007 20:04 > To: Andy Grove > Cc: rhm-users at redhat.com > Subject: Re: [Rhm-users] Which version of RHM C++ broker is compatible with Apache QPID M2? > > Andy, > > The M2 release is based on the AMQP 0-8 spec and was done based on the > M2 branch. > > RHM c++ broker and the clients are based on the 0-10 version of the AMQP > spec (which will be released shortly). This is based on the Apache Qpid > trunk. > > Along with the RHM c++ broker we have clients in c++, java (jms) and > python that will work this broker. > > Regards, > > Rajith > > > On Thu, 2007-12-06 at 19:48 +0000, Andy Grove wrote: > >> Hi, >> >> >> >> I?ve been using Apache QPID for a while now as part of a solution I?m >> working on and I would like to be able to start testing the RHM C++ >> broker instead of the Apache QPID Java broker. Previously when I tried >> this there was a compatibility issue between the QPID client and the >> RHM broker. Is there now a version of the RHM C++ broker that is >> compatible with the recent QPID M2 release? >> >> >> >> Thanks, >> >> >> >> Andy. >> >> >> _______________________________________________ >> Rhm-users mailing list >> Rhm-users at redhat.com >> https://www.redhat.com/mailman/listinfo/rhm-users >> > > > _______________________________________________ > Rhm-users mailing list > Rhm-users at redhat.com > https://www.redhat.com/mailman/listinfo/rhm-users > -------------- next part -------------- An HTML attachment was scrubbed... URL: From andy at codesuccess.com Thu Dec 6 21:32:40 2007 From: andy at codesuccess.com (Andy Grove) Date: Thu, 6 Dec 2007 21:32:40 -0000 Subject: [Rhm-users] Which version of RHM C++ broker is compatible with Apache QPID M2? In-Reply-To: <475866A6.4040902@redhat.com> References: <028201c83840$fd0959d0$f71c0d70$@com> <1196971421.4135.46.camel@localhost.localdomain> <02ae01c8384b$ceb28dd0$6c17a970$@com> <475866A6.4040902@redhat.com> Message-ID: <02b201c8384f$88378190$98a684b0$@com> Thanks Carl. I look forward to trying this out in the morning (it?s getting late here in the UK). Please do send me the yum repo details when you have them. I have time tomorrow to run some benchmarks and I would be happy to share the results. Andy. From: Carl Trieloff [mailto:cctrieloff at redhat.com] Sent: 06 December 2007 21:16 To: Andy Grove Cc: rattapat at redhat.com; rhm-users at redhat.com Subject: Re: [Rhm-users] Which version of RHM C++ broker is compatible with Apache QPID M2? Full instructions to do it all from rpm's should sync live a bit later today. A download page will appear on www.redhat.com/mrg We can also post you the yum repo URL once the sync up is complete. Carl. Andy Grove wrote: Thanks for the quick response. What's the quickest way to get up and running with a Java client? Do I need to check the source out of subversion and build it or should I use one of the source RPMs that are available for download? Are there any binary distributions for the client? Thanks, Andy. -----Original Message----- From: Rajith Attapattu [mailto:rattapat at redhat.com] Sent: 06 December 2007 20:04 To: Andy Grove Cc: rhm-users at redhat.com Subject: Re: [Rhm-users] Which version of RHM C++ broker is compatible with Apache QPID M2? Andy, The M2 release is based on the AMQP 0-8 spec and was done based on the M2 branch. RHM c++ broker and the clients are based on the 0-10 version of the AMQP spec (which will be released shortly). This is based on the Apache Qpid trunk. Along with the RHM c++ broker we have clients in c++, java (jms) and python that will work this broker. Regards, Rajith On Thu, 2007-12-06 at 19:48 +0000, Andy Grove wrote: Hi, I?ve been using Apache QPID for a while now as part of a solution I?m working on and I would like to be able to start testing the RHM C++ broker instead of the Apache QPID Java broker. Previously when I tried this there was a compatibility issue between the QPID client and the RHM broker. Is there now a version of the RHM C++ broker that is compatible with the recent QPID M2 release? Thanks, Andy. _______________________________________________ Rhm-users mailing list Rhm-users at redhat.com https://www.redhat.com/mailman/listinfo/rhm-users _______________________________________________ Rhm-users mailing list Rhm-users at redhat.com https://www.redhat.com/mailman/listinfo/rhm-users -------------- next part -------------- An HTML attachment was scrubbed... URL: From andy at codesuccess.com Fri Dec 7 07:51:07 2007 From: andy at codesuccess.com (Andy Grove) Date: Fri, 7 Dec 2007 07:51:07 +0000 Subject: [Rhm-users] Problem installing MRG beta 1 Message-ID: <8e5324b0712062351s46badb39q1e23a1cc061a3cdf@mail.gmail.com> I've just tried to install rhm using yum as per the instructions on the MRG site and I've run into an installation problem. I am missing these dependencies: Error: Missing Dependency: libboost_program_options.so.2()(64bit) is needed by package qpidd Error: Missing Dependency: libboost_iostreams.so.2()(64bit) is needed by package qpidd Error: Missing Dependency: libboost_program_options.so.2()(64bit) is needed by package qpidc However, if I run "yum install boost" I get this output: Package boost - 1.34.1-5.fc8.i386 is already installed. Package boost - 1.34.1-5.fc8.x86_64 is already installed. Nothing to do I also have boost-devel and boost-devel-static installed. I'm by no means an expert when it comes to installing software on linux so I expect I'm missing something obvious. I'm running Fedora 8 on Intel hardware. Any assistance would be appreciated. Thanks, Andy. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gsim at redhat.com Fri Dec 7 08:56:19 2007 From: gsim at redhat.com (Gordon Sim) Date: Fri, 07 Dec 2007 08:56:19 +0000 Subject: [Rhm-users] Problem installing MRG beta 1 In-Reply-To: <8e5324b0712062351s46badb39q1e23a1cc061a3cdf@mail.gmail.com> References: <8e5324b0712062351s46badb39q1e23a1cc061a3cdf@mail.gmail.com> Message-ID: <47590AB3.4000009@redhat.com> Andy Grove wrote: > I've just tried to install rhm using yum as per the instructions on the > MRG site and I've run into an installation problem. I am missing these > dependencies: > > Error: Missing Dependency: libboost_program_options.so.2()(64bit) is > needed by package qpidd > Error: Missing Dependency: libboost_iostreams.so.2()(64bit) is needed by > package qpidd > Error: Missing Dependency: libboost_program_options.so.2()(64bit) is > needed by package qpidc > > However, if I run "yum install boost" I get this output: > > Package boost - 1.34.1-5.fc8.i386 is already installed. > Package boost - 1.34.1-5.fc8.x86_64 is already installed. > Nothing to do > > I also have boost-devel and boost-devel-static installed. I'm by no > means an expert when it comes to installing software on linux so I > expect I'm missing something obvious. I'm running Fedora 8 on Intel > hardware. > > Any assistance would be appreciated. Andy, The packages in that repo are for RHEL5. Would you mind building from source? To do so, download http://rhm.et.redhat.com/download/qpidc-0.2.tar.gz, untar it and in the resulting qpidc directory run ./configure then make and make install. The persistence module is also available and the same process should work after you have installed qpidc: http://rhm.et.redhat.com/download/rhm-0.2.tar.gz I hope this helps, --Gordon. From andy at codesuccess.com Fri Dec 7 09:43:07 2007 From: andy at codesuccess.com (Andy Grove) Date: Fri, 7 Dec 2007 09:43:07 +0000 Subject: [Rhm-users] Problem installing MRG beta 1 In-Reply-To: <47590AB3.4000009@redhat.com> References: <8e5324b0712062351s46badb39q1e23a1cc061a3cdf@mail.gmail.com> <47590AB3.4000009@redhat.com> Message-ID: <8e5324b0712070143o7abfc9f6h2f0c188a9e4491ef@mail.gmail.com> I tried compiling from source using the tar from the link you supplied but I get some compilation errors - here is the first one: qpid/broker/SemanticState.cpp: In destructor 'virtual qpid::broker::SemanticState::~SemanticState()': qpid/broker/SemanticState.cpp:74: error: no matching function for call to 'qpid::broker::SemanticState::cancel(boost::ptr_container_detail::ref_pair, std::allocator >, qpid::broker::SemanticState::ConsumerImpl* const>)' qpid/broker/SemanticState.h:129: note: candidates are: void qpid::broker::SemanticState::cancel(qpid::broker::SemanticState::ConsumerImpl&) qpid/broker/SemanticState.h:156: note: void qpid::broker::SemanticState::cancel(const std::string&) Any idea why I am getting these errors? Is it because I'm using Fedora 8 rather than RHEL5? Thanks, Andy. On 12/7/07, Gordon Sim wrote: > > Andy Grove wrote: > > I've just tried to install rhm using yum as per the instructions on the > > MRG site and I've run into an installation problem. I am missing these > > dependencies: > > > > Error: Missing Dependency: libboost_program_options.so.2()(64bit) is > > needed by package qpidd > > Error: Missing Dependency: libboost_iostreams.so.2()(64bit) is needed by > > package qpidd > > Error: Missing Dependency: libboost_program_options.so.2()(64bit) is > > needed by package qpidc > > > > However, if I run "yum install boost" I get this output: > > > > Package boost - 1.34.1-5.fc8.i386 is already installed. > > Package boost - 1.34.1-5.fc8.x86_64 is already installed. > > Nothing to do > > > > I also have boost-devel and boost-devel-static installed. I'm by no > > means an expert when it comes to installing software on linux so I > > expect I'm missing something obvious. I'm running Fedora 8 on Intel > > hardware. > > > > Any assistance would be appreciated. > > Andy, > > The packages in that repo are for RHEL5. Would you mind building from > source? > > To do so, download http://rhm.et.redhat.com/download/qpidc-0.2.tar.gz, > untar it and in the resulting qpidc directory run ./configure then make > and make install. > > The persistence module is also available and the same process should > work after you have installed qpidc: > http://rhm.et.redhat.com/download/rhm-0.2.tar.gz > > I hope this helps, > > --Gordon. > -------------- next part -------------- An HTML attachment was scrubbed... URL: From gsim at redhat.com Fri Dec 7 10:00:47 2007 From: gsim at redhat.com (Gordon Sim) Date: Fri, 07 Dec 2007 10:00:47 +0000 Subject: [Rhm-users] Problem installing MRG beta 1 In-Reply-To: <8e5324b0712070143o7abfc9f6h2f0c188a9e4491ef@mail.gmail.com> References: <8e5324b0712062351s46badb39q1e23a1cc061a3cdf@mail.gmail.com> <47590AB3.4000009@redhat.com> <8e5324b0712070143o7abfc9f6h2f0c188a9e4491ef@mail.gmail.com> Message-ID: <475919CF.7050608@redhat.com> Andy Grove wrote: > I tried compiling from source using the tar from the link you supplied > but I get some compilation errors - here is the first one: > > qpid/broker/SemanticState.cpp: In destructor 'virtual > qpid::broker::SemanticState::~SemanticState()': > qpid/broker/SemanticState.cpp:74: error: no matching function for call > to > 'qpid::broker::SemanticState::cancel(boost::ptr_container_detail::ref_pair std::char_traits, std::allocator >, > qpid::broker::SemanticState::ConsumerImpl* const>)' > qpid/broker/SemanticState.h:129: note: candidates are: void > qpid::broker::SemanticState::cancel(qpid::broker::SemanticState::ConsumerImpl&) > qpid/broker/SemanticState.h:156: note: void > qpid::broker::SemanticState::cancel(const std::string&) > > Any idea why I am getting these errors? Is it because I'm using Fedora 8 > rather than RHEL5? Sorry Andy, that looks like an issue with the boost version which is indeed different on Fedora 8. We do know about this and I thought we had it all fixed, but clearly not. I'll look into this and we can hopefully get you an update that fixes this. Apologies again and thanks for your patience! From astitcher at redhat.com Fri Dec 7 11:35:44 2007 From: astitcher at redhat.com (Andrew Stitcher) Date: Fri, 07 Dec 2007 11:35:44 +0000 Subject: [Rhm-users] Problem installing MRG beta 1 In-Reply-To: <8e5324b0712070143o7abfc9f6h2f0c188a9e4491ef@mail.gmail.com> References: <8e5324b0712062351s46badb39q1e23a1cc061a3cdf@mail.gmail.com> <47590AB3.4000009@redhat.com> <8e5324b0712070143o7abfc9f6h2f0c188a9e4491ef@mail.gmail.com> Message-ID: <1197027344.3481.5.camel@fuschia.home> On Fri, 2007-12-07 at 09:43 +0000, Andy Grove wrote: > I tried compiling from source using the tar from the link you supplied > but I get some compilation errors - here is the first one: > > qpid/broker/SemanticState.cpp: In destructor 'virtual > qpid::broker::SemanticState::~SemanticState()': > qpid/broker/SemanticState.cpp:74: error: no matching function for call > to > 'qpid::broker::SemanticState::cancel(boost::ptr_container_detail::ref_pair, std::allocator >, qpid::broker::SemanticState::ConsumerImpl* const>)' > qpid/broker/SemanticState.h:129: note: candidates are: void > qpid::broker::SemanticState::cancel(qpid::broker::SemanticState::ConsumerImpl&) > qpid/broker/SemanticState.h:156: note: void > qpid::broker::SemanticState::cancel(const std::string&) > > Any idea why I am getting these errors? Is it because I'm using Fedora > 8 rather than RHEL5? I know they were some problems compiling the qpid C++ broker with F8, and I'm not sure whether they were all fixed. I don't think F8 is a regular development platform for any of the main C++ broker contributors yet, so this could have slipped in. BTW it appears that the cause is a different version of boost in F8, so downgrading it to the version in F7 would probably fix this issue. Andrew From gsim at redhat.com Fri Dec 7 12:22:24 2007 From: gsim at redhat.com (Gordon Sim) Date: Fri, 07 Dec 2007 12:22:24 +0000 Subject: [Rhm-users] Problem installing MRG beta 1 In-Reply-To: <1197027344.3481.5.camel@fuschia.home> References: <8e5324b0712062351s46badb39q1e23a1cc061a3cdf@mail.gmail.com> <47590AB3.4000009@redhat.com> <8e5324b0712070143o7abfc9f6h2f0c188a9e4491ef@mail.gmail.com> <1197027344.3481.5.camel@fuschia.home> Message-ID: <47593B00.7070208@redhat.com> Andrew Stitcher wrote: > I know they were some problems compiling the qpid C++ broker with F8, > and I'm not sure whether they were all fixed. I don't think F8 is a > regular development platform for any of the main C++ broker contributors > yet, so this could have slipped in. BTW it appears that the cause is a > different version of boost in F8, so downgrading it to the version in F7 > would probably fix this issue. I have fixed the specific issue reported, I believe, I'm just waiting for verification that everything builds ok on Fedora 8 (which I don't have to hand). We can then make a newer version of the source release available. From andy at codesuccess.com Fri Dec 7 12:53:59 2007 From: andy at codesuccess.com (Andy Grove) Date: Fri, 7 Dec 2007 12:53:59 -0000 Subject: [Rhm-users] Problem installing MRG beta 1 In-Reply-To: <1197027344.3481.5.camel@fuschia.home> References: <8e5324b0712062351s46badb39q1e23a1cc061a3cdf@mail.gmail.com> <47590AB3.4000009@redhat.com> <8e5324b0712070143o7abfc9f6h2f0c188a9e4491ef@mail.gmail.com> <1197027344.3481.5.camel@fuschia.home> Message-ID: <001601c838d0$3d97a300$b8c6e900$@com> Andrew, Thanks for the suggestion. Excuse my ignorance but how exactly do I downgrade boost. I assume I need to "yum remove boost" and then "yum install boost-n.n.n.fc7" ? Can you tell me exactly which version I need or how to get a list of available versions? Thanks, Andy. -----Original Message----- From: Andrew Stitcher [mailto:astitcher at redhat.com] Sent: 07 December 2007 11:36 To: Andy Grove Cc: Gordon Sim; rhm-users at redhat.com Subject: Re: [Rhm-users] Problem installing MRG beta 1 On Fri, 2007-12-07 at 09:43 +0000, Andy Grove wrote: > I tried compiling from source using the tar from the link you supplied > but I get some compilation errors - here is the first one: > > qpid/broker/SemanticState.cpp: In destructor 'virtual > qpid::broker::SemanticState::~SemanticState()': > qpid/broker/SemanticState.cpp:74: error: no matching function for call > to > 'qpid::broker::SemanticState::cancel(boost::ptr_container_detail::ref_pair, std::allocator >, qpid::broker::SemanticState::ConsumerImpl* const>)' > qpid/broker/SemanticState.h:129: note: candidates are: void > qpid::broker::SemanticState::cancel(qpid::broker::SemanticState::ConsumerImp l&) > qpid/broker/SemanticState.h:156: note: void > qpid::broker::SemanticState::cancel(const std::string&) > > Any idea why I am getting these errors? Is it because I'm using Fedora > 8 rather than RHEL5? I know they were some problems compiling the qpid C++ broker with F8, and I'm not sure whether they were all fixed. I don't think F8 is a regular development platform for any of the main C++ broker contributors yet, so this could have slipped in. BTW it appears that the cause is a different version of boost in F8, so downgrading it to the version in F7 would probably fix this issue. Andrew From andy at codesuccess.com Fri Dec 7 12:54:29 2007 From: andy at codesuccess.com (Andy Grove) Date: Fri, 7 Dec 2007 12:54:29 -0000 Subject: [Rhm-users] Problem installing MRG beta 1 In-Reply-To: <47593B00.7070208@redhat.com> References: <8e5324b0712062351s46badb39q1e23a1cc061a3cdf@mail.gmail.com> <47590AB3.4000009@redhat.com> <8e5324b0712070143o7abfc9f6h2f0c188a9e4491ef@mail.gmail.com> <1197027344.3481.5.camel@fuschia.home> <47593B00.7070208@redhat.com> Message-ID: <001701c838d0$4f1b4780$ed51d680$@com> That's great news. Thanks, Gordon. I'll be ready to try this as soon as it is available. Andy. -----Original Message----- From: Gordon Sim [mailto:gsim at redhat.com] Sent: 07 December 2007 12:22 To: Andrew Stitcher Cc: Andy Grove; rhm-users at redhat.com Subject: Re: [Rhm-users] Problem installing MRG beta 1 Andrew Stitcher wrote: > I know they were some problems compiling the qpid C++ broker with F8, > and I'm not sure whether they were all fixed. I don't think F8 is a > regular development platform for any of the main C++ broker contributors > yet, so this could have slipped in. BTW it appears that the cause is a > different version of boost in F8, so downgrading it to the version in F7 > would probably fix this issue. I have fixed the specific issue reported, I believe, I'm just waiting for verification that everything builds ok on Fedora 8 (which I don't have to hand). We can then make a newer version of the source release available. From astitcher at redhat.com Fri Dec 7 13:02:28 2007 From: astitcher at redhat.com (Andrew Stitcher) Date: Fri, 07 Dec 2007 13:02:28 +0000 Subject: [Rhm-users] Problem installing MRG beta 1 In-Reply-To: <001601c838d0$3d97a300$b8c6e900$@com> References: <8e5324b0712062351s46badb39q1e23a1cc061a3cdf@mail.gmail.com> <47590AB3.4000009@redhat.com> <8e5324b0712070143o7abfc9f6h2f0c188a9e4491ef@mail.gmail.com> <1197027344.3481.5.camel@fuschia.home> <001601c838d0$3d97a300$b8c6e900$@com> Message-ID: <1197032548.3481.12.camel@fuschia.home> On Fri, 2007-12-07 at 12:53 +0000, Andy Grove wrote: > Andrew, > > Thanks for the suggestion. Excuse my ignorance but how exactly do I > downgrade boost. I assume I need to "yum remove boost" and then "yum install > boost-n.n.n.fc7" ? Can you tell me exactly which version I need or how to > get a list of available versions? You'll probably need to get the boost pkgs individually from the F7 repo. The versions I have here are: boost-1.33.1-13.fc7.x86_64 boost-1.33.1-13.fc7.i386 I'd try googling directly for boost-1.33.1-13.fc7. and download the rpm. then you can use rpm -Uhv ... to install the rpm directly. If you're not desperate it'll probably be easier to just wait for us to fix the src and re-release. Andrew > Thanks, > > Andy. > > -----Original Message----- > From: Andrew Stitcher [mailto:astitcher at redhat.com] > Sent: 07 December 2007 11:36 > To: Andy Grove > Cc: Gordon Sim; rhm-users at redhat.com > Subject: Re: [Rhm-users] Problem installing MRG beta 1 > > On Fri, 2007-12-07 at 09:43 +0000, Andy Grove wrote: > > I tried compiling from source using the tar from the link you supplied > > but I get some compilation errors - here is the first one: > > > > qpid/broker/SemanticState.cpp: In destructor 'virtual > > qpid::broker::SemanticState::~SemanticState()': > > qpid/broker/SemanticState.cpp:74: error: no matching function for call > > to > > > 'qpid::broker::SemanticState::cancel(boost::ptr_container_detail::ref_pair td::basic_string, std::allocator >, > qpid::broker::SemanticState::ConsumerImpl* const>)' > > qpid/broker/SemanticState.h:129: note: candidates are: void > > > qpid::broker::SemanticState::cancel(qpid::broker::SemanticState::ConsumerImp > l&) > > qpid/broker/SemanticState.h:156: note: void > > qpid::broker::SemanticState::cancel(const std::string&) > > > > Any idea why I am getting these errors? Is it because I'm using Fedora > > 8 rather than RHEL5? > > I know they were some problems compiling the qpid C++ broker with F8, > and I'm not sure whether they were all fixed. I don't think F8 is a > regular development platform for any of the main C++ broker contributors > yet, so this could have slipped in. BTW it appears that the cause is a > different version of boost in F8, so downgrading it to the version in F7 > would probably fix this issue. > > Andrew > > From gsim at redhat.com Fri Dec 7 14:35:38 2007 From: gsim at redhat.com (Gordon Sim) Date: Fri, 07 Dec 2007 14:35:38 +0000 Subject: [Rhm-users] Problem installing MRG beta 1 In-Reply-To: <001701c838d0$4f1b4780$ed51d680$@com> References: <8e5324b0712062351s46badb39q1e23a1cc061a3cdf@mail.gmail.com> <47590AB3.4000009@redhat.com> <8e5324b0712070143o7abfc9f6h2f0c188a9e4491ef@mail.gmail.com> <1197027344.3481.5.camel@fuschia.home> <47593B00.7070208@redhat.com> <001701c838d0$4f1b4780$ed51d680$@com> Message-ID: <47595A3A.5040306@redhat.com> Andy, There is a new tar that should build on fedora 8 now if you'd like to try it: http://rhm.et.redhat.com/download/qpidc-0.2.tar.gz Thanks again, --Gordon. From andy at codesuccess.com Fri Dec 7 14:53:39 2007 From: andy at codesuccess.com (Andy Grove) Date: Fri, 7 Dec 2007 14:53:39 -0000 Subject: [Rhm-users] Problem installing MRG beta 1 In-Reply-To: <47595A3A.5040306@redhat.com> References: <8e5324b0712062351s46badb39q1e23a1cc061a3cdf@mail.gmail.com> <47590AB3.4000009@redhat.com> <8e5324b0712070143o7abfc9f6h2f0c188a9e4491ef@mail.gmail.com> <1197027344.3481.5.camel@fuschia.home> <47593B00.7070208@redhat.com> <001701c838d0$4f1b4780$ed51d680$@com> <47595A3A.5040306@redhat.com> Message-ID: <001801c838e0$f52e9680$df8bc380$@com> Gordon, Many Thanks. That compiles cleanly. I'm going to try testing a java client against the broker now. Andy. -----Original Message----- From: Gordon Sim [mailto:gsim at redhat.com] Sent: 07 December 2007 14:36 To: Andy Grove Cc: rhm-users at redhat.com Subject: Re: [Rhm-users] Problem installing MRG beta 1 Andy, There is a new tar that should build on fedora 8 now if you'd like to try it: http://rhm.et.redhat.com/download/qpidc-0.2.tar.gz Thanks again, --Gordon. From andy at codesuccess.com Sat Dec 8 08:43:23 2007 From: andy at codesuccess.com (Andy Grove) Date: Sat, 8 Dec 2007 08:43:23 +0000 Subject: [Rhm-users] Unable to connect using Java client with RHM beta Message-ID: <8e5324b0712080043k6bbefe39x23e0b11d709c7abe@mail.gmail.com> Hi, I installed the RHM java client version 0.2.4 from RPMs yesterday and I am trying to connect to the RHM C++ broker but connections timeout. If I kill the RHM broker and run the Apache QPID Java broker the client connects fine. Do I need to change my connection settings for RHM? These are the settings I am currently using: java.naming.factory.initial = org.apache.qpid.jndi.PropertiesFileInitialContextFactory connectionfactory.local = amqp://guest:guest at clientid /test?brokerlist='tcp://localhost:5672' I'm using the following Java code to connect: InitialContext ctx = new InitialContext(properties); AMQConnectionFactory cf = (AMQConnectionFactory) ctx.lookup("local"); Connection conn = cf.createConnection(); Thanks, Andy. -------------- next part -------------- An HTML attachment was scrubbed... URL: From andy at codesuccess.com Sat Dec 8 09:08:21 2007 From: andy at codesuccess.com (Andy Grove) Date: Sat, 8 Dec 2007 09:08:21 -0000 Subject: [Rhm-users] RE: Unable to connect using Java client with RHM beta In-Reply-To: <8e5324b0712080043k6bbefe39x23e0b11d709c7abe@mail.gmail.com> References: <8e5324b0712080043k6bbefe39x23e0b11d709c7abe@mail.gmail.com> Message-ID: <000301c83979$e2aaa780$a7fff680$@com> After reading the documentation I realized I need to change the connection details. I think I have everything working now. Thanks, Andy. From: Andy Grove [mailto:andy at codesuccess.com] Sent: 08 December 2007 08:43 To: rhm-users at redhat.com Subject: Unable to connect using Java client with RHM beta Hi, I installed the RHM java client version 0.2.4 from RPMs yesterday and I am trying to connect to the RHM C++ broker but connections timeout. If I kill the RHM broker and run the Apache QPID Java broker the client connects fine. Do I need to change my connection settings for RHM? These are the settings I am currently using: java.naming.factory.initial = org.apache.qpid.jndi.PropertiesFileInitialContextFactory connectionfactory.local = amqp://guest:guest at clientid/test?brokerlist='tcp://localhost:5672' I'm using the following Java code to connect: InitialContext ctx = new InitialContext(properties); AMQConnectionFactory cf = (AMQConnectionFactory) ctx.lookup("local"); Connection conn = cf.createConnection(); Thanks, Andy. -------------- next part -------------- An HTML attachment was scrubbed... URL: From andy at codesuccess.com Sat Dec 8 10:20:13 2007 From: andy at codesuccess.com (Andy Grove) Date: Sat, 8 Dec 2007 10:20:13 -0000 Subject: [Rhm-users] RHM performance Message-ID: <001a01c83983$ecbb8370$c6328a50$@com> Hi I've just been running some benchmarks against the RHM broker and the Apache QPID Java broker. They are showing very different performance and scalability characteristics and I'm rather surprised to see that the RHM C++ broker is many times slower than the Apache QPID Java broker when running these tests. I'm hoping that this is just a configuration issue. I've posted a graph of the results in PDF format at the following location: http://www.codesuccess.com/temp/rhm_qpid_perf.pdf Can anyone shed any light on these results? I was assuming that performance would be better with the RHM C++ broker. Here's a brief description of the tests: The test consists of client and a server. The server creates a durable subscriber for 50 topics and listens for messages (non transactional, auto acknowledge). The client that starts with a single thread and then adds a new thread every 5 seconds. Each thread continuously publishes a fixed size message to one of 50 topics at random. Throughput is measured on the client for each 5 second duration (the cost of setting up each thread and connecting to JMS is excluded from the throughput measurement). The client uses persistent messaging and transactions. The test client, test server and broker are each running on a separate quad core 2.4GHz Intel PC with 4GBM RAM running Fedora 8 (64 bit). Each machine has an identical hardware specification. The PCs are all connected to the same gigabit switch. I used the same client java jars (from the RHM RPM) when testing against the RHM C++ broker and the Apache QPID Java broker. The RHM broker was not configured with a persistent store but QPID was. RHM and QPID brokers were run on the same server, but not at the same time. Thanks, Andy. -------------- next part -------------- An HTML attachment was scrubbed... URL: From gsim at redhat.com Mon Dec 10 08:09:02 2007 From: gsim at redhat.com (Gordon Sim) Date: Mon, 10 Dec 2007 08:09:02 +0000 Subject: [Rhm-users] RHM performance In-Reply-To: <001a01c83983$ecbb8370$c6328a50$@com> References: <001a01c83983$ecbb8370$c6328a50$@com> Message-ID: <475CF41E.3020102@redhat.com> Andy Grove wrote: > I?ve just been running some benchmarks against the RHM broker and the > Apache QPID Java broker. They are showing very different performance and > scalability characteristics and I?m rather surprised to see that the RHM > C++ broker is many times slower than the Apache QPID Java broker when > running these tests. I?m hoping that this is just a configuration > issue. Since you don't have persistence modules loaded for each broker[*], it may be a configuration issue in the client (though the jars are the same the actual classes used are different depending on which version of the protocol is used). The broker should support much higher throughput than you are seeing and there is a known issue with the JMS performance for this beta. I'll get some more detail on this for you. One question, are you using a MessageListener or receive()? [*] if and when you do install persistence for the c++ broker, the default mode is also very slow, this can be improved significantly be specifying '--store-async yes' to use aio. From cctrieloff at redhat.com Mon Dec 10 14:28:39 2007 From: cctrieloff at redhat.com (Carl Trieloff) Date: Mon, 10 Dec 2007 09:28:39 -0500 Subject: [Rhm-users] RHM performance In-Reply-To: <001a01c83983$ecbb8370$c6328a50$@com> References: <001a01c83983$ecbb8370$c6328a50$@com> Message-ID: <475D4D17.7070302@redhat.com> > > > > > The RHM broker was not configured with a persistent store but QPID > was. RHM and QPID brokers were run on the same server, but not at the > same time. > > > Andy, What persistent store options did you use? ( as Gordon points out, use --store-async yes and to clear the previous config use --store-force yes (this will clear the previous data) Something else is amiss which we need to figure out. Carl. -------------- next part -------------- An HTML attachment was scrubbed... URL: From aconway at redhat.com Mon Dec 10 14:25:26 2007 From: aconway at redhat.com (Alan Conway) Date: Mon, 10 Dec 2007 09:25:26 -0500 Subject: [Rhm-users] RHM performance In-Reply-To: <001a01c83983$ecbb8370$c6328a50$@com> References: <001a01c83983$ecbb8370$c6328a50$@com> Message-ID: <475D4C56.5060505@redhat.com> Andy Grove wrote: > Hi > > I?ve just been running some benchmarks against the RHM broker and the > Apache QPID Java broker. They are showing very different performance and > scalability characteristics and I?m rather surprised to see that the RHM > C++ broker is many times slower than the Apache QPID Java broker when > running these tests. I?m hoping that this is just a configuration > issue. I?ve posted a graph of the results in PDF format at the following > location: > > http://www.codesuccess.com/temp/rhm_qpid_perf.pdf > > Can anyone shed any light on these results? I was assuming that > performance would be better with the RHM C++ broker. I'd like to investigate this a little. Any chance you could send me your test code? If not, no problem I can follow your description. Also can you give me some details on the machine you used for the tests? In particular if you could attach /proc/cpuinfo and /proc/meminfo. Thanks, Alan. From aconway at redhat.com Mon Dec 10 15:47:00 2007 From: aconway at redhat.com (Alan Conway) Date: Mon, 10 Dec 2007 10:47:00 -0500 Subject: [Rhm-users] RHM performance In-Reply-To: <001a01c83983$ecbb8370$c6328a50$@com> References: <001a01c83983$ecbb8370$c6328a50$@com> Message-ID: <475D5F74.9030000@redhat.com> I reread your message and I see you already told me the machine specs. Can you tell me what size of message you used? Andy Grove wrote: > Hi > > > > I?ve just been running some benchmarks against the RHM broker and the > Apache QPID Java broker. They are showing very different performance and > scalability characteristics and I?m rather surprised to see that the RHM > C++ broker is many times slower than the Apache QPID Java broker when > running these tests. I?m hoping that this is just a configuration > issue. I?ve posted a graph of the results in PDF format at the following > location: > > > > http://www.codesuccess.com/temp/rhm_qpid_perf.pdf > > > > Can anyone shed any light on these results? I was assuming that > performance would be better with the RHM C++ broker. > > > > Here?s a brief description of the tests: > > > > The test consists of client and a server. > > > > The server creates a durable subscriber for 50 topics and listens for > messages (non transactional, auto acknowledge). > > > > The client that starts with a single thread and then adds a new thread > every 5 seconds. Each thread continuously publishes a fixed size message > to one of 50 topics at random. Throughput is measured on the client for > each 5 second duration (the cost of setting up each thread and > connecting to JMS is excluded from the throughput measurement). The > client uses persistent messaging and transactions. > > > > The test client, test server and broker are each running on a separate > quad core 2.4GHz Intel PC with 4GBM RAM running Fedora 8 (64 bit). Each > machine has an identical hardware specification. The PCs are all > connected to the same gigabit switch. > > > > I used the same client java jars (from the RHM RPM) when testing against > the RHM C++ broker and the Apache QPID Java broker. > > > > The RHM broker was not configured with a persistent store but QPID was. > RHM and QPID brokers were run on the same server, but not at the same time. > > > > Thanks, > > > > Andy. > > > > > ------------------------------------------------------------------------ > > _______________________________________________ > Rhm-users mailing list > Rhm-users at redhat.com > https://www.redhat.com/mailman/listinfo/rhm-users From andy at codesuccess.com Mon Dec 10 17:20:38 2007 From: andy at codesuccess.com (=?utf-8?B?QW5keSBHcm92ZQ==?=) Date: Mon, 10 Dec 2007 17:20:38 +0000 Subject: [Rhm-users] RHM performance In-Reply-To: <475D5F74.9030000@redhat.com> References: <001a01c83983$ecbb8370$c6328a50$@com><475D5F74.9030000@redhat.com> Message-ID: <2120630074-1197307253-cardhu_decombobulator_blackberry.rim.net-428748744-@bxe013.bisx.produk.on.blackberry> Hi Alan, I'm away for a few days on business but I'm pretty sure I was using 1024 byte messages for these.tests. I think it would be useful if I made the source available when I get back to the UK at the end of the week. Thanks, Andy. Sent from my BlackBerry? wireless device -----Original Message----- From: Alan Conway Date: Mon, 10 Dec 2007 10:47:00 To:Andy Grove Cc:rhm-users at redhat.com Subject: Re: [Rhm-users] RHM performance I reread your message and I see you already told me the machine specs. Can you tell me what size of message you used? Andy Grove wrote: > Hi > > > > I?ve just been running some benchmarks against the RHM broker and the > Apache QPID Java broker. They are showing very different performance and > scalability characteristics and I?m rather surprised to see that the RHM > C++ broker is many times slower than the Apache QPID Java broker when > running these tests. I?m hoping that this is just a configuration > issue. I?ve posted a graph of the results in PDF format at the following > location: > > > > http://www.codesuccess.com/temp/rhm_qpid_perf.pdf > > > > Can anyone shed any light on these results? I was assuming that > performance would be better with the RHM C++ broker. > > > > Here?s a brief description of the tests: > > > > The test consists of client and a server. > > > > The server creates a durable subscriber for 50 topics and listens for > messages (non transactional, auto acknowledge). > > > > The client that starts with a single thread and then adds a new thread > every 5 seconds. Each thread continuously publishes a fixed size message > to one of 50 topics at random. Throughput is measured on the client for > each 5 second duration (the cost of setting up each thread and > connecting to JMS is excluded from the throughput measurement). The > client uses persistent messaging and transactions. > > > > The test client, test server and broker are each running on a separate > quad core 2.4GHz Intel PC with 4GBM RAM running Fedora 8 (64 bit). Each > machine has an identical hardware specification. The PCs are all > connected to the same gigabit switch. > > > > I used the same client java jars (from the RHM RPM) when testing against > the RHM C++ broker and the Apache QPID Java broker. > > > > The RHM broker was not configured with a persistent store but QPID was. > RHM and QPID brokers were run on the same server, but not at the same time. > > > > Thanks, > > > > Andy. > > > > > ------------------------------------------------------------------------ > > _______________________________________________ > Rhm-users mailing list > Rhm-users at redhat.com > https://www.redhat.com/mailman/listinfo/rhm-users From gsim at redhat.com Mon Dec 10 19:34:10 2007 From: gsim at redhat.com (Gordon Sim) Date: Mon, 10 Dec 2007 19:34:10 +0000 Subject: [Rhm-users] RHM performance In-Reply-To: <001a01c83983$ecbb8370$c6328a50$@com> References: <001a01c83983$ecbb8370$c6328a50$@com> Message-ID: <475D94B2.3090205@redhat.com> Andy Grove wrote: > The client that starts with a single thread and then adds a new thread > every 5 seconds. Each thread continuously publishes a fixed size message > to one of 50 topics at random. Throughput is measured on the client for > each 5 second duration (the cost of setting up each thread and > connecting to JMS is excluded from the throughput measurement). I suspect a large part of the difference between the measurements for the two systems may be attributable to the fact that publish is synchronous for durable messages in 0-10 (the next message is not sent until the broker acknowledges the previous one) whereas in 0-8 there is no confirmation of the publication (the fact that send() returns does not necessarily imply that the broker has even received it, let alone stored it). From gsim at redhat.com Tue Dec 11 08:01:24 2007 From: gsim at redhat.com (Gordon Sim) Date: Tue, 11 Dec 2007 08:01:24 +0000 Subject: [Rhm-users] RHM performance In-Reply-To: <475D94B2.3090205@redhat.com> References: <001a01c83983$ecbb8370$c6328a50$@com> <475D94B2.3090205@redhat.com> Message-ID: <475E43D4.8020008@redhat.com> Gordon Sim wrote: > Andy Grove wrote: >> The client that starts with a single thread and then adds a new thread >> every 5 seconds. Each thread continuously publishes a fixed size >> message to one of 50 topics at random. Throughput is measured on the >> client for each 5 second duration (the cost of setting up each thread >> and connecting to JMS is excluded from the throughput measurement). > > I suspect a large part of the difference between the measurements for > the two systems may be attributable to the fact that publish is > synchronous for durable messages in 0-10 (the next message is not sent > until the broker acknowledges the previous one) whereas in 0-8 there is > no confirmation of the publication (the fact that send() returns does > not necessarily imply that the broker has even received it, let alone > stored it). Sorry, I may have inadvertently given the impression that the 0-10 protocol requires synchronous publish for durable messages. That is not the case. It is an implementation in the current jms client and will likely soon be configurable. From aconway at redhat.com Tue Dec 11 13:27:00 2007 From: aconway at redhat.com (Alan Conway) Date: Tue, 11 Dec 2007 08:27:00 -0500 Subject: [Rhm-users] RHM performance In-Reply-To: <2120630074-1197307253-cardhu_decombobulator_blackberry.rim.net-428748744-@bxe013.bisx.produk.on.blackberry> References: <001a01c83983$ecbb8370$c6328a50$@com><475D5F74.9030000@redhat.com> <2120630074-1197307253-cardhu_decombobulator_blackberry.rim.net-428748744-@bxe013.bisx.produk.on.blackberry> Message-ID: <475E9024.9050208@redhat.com> Andy Grove wrote: > Hi Alan, > > I'm away for a few days on business but I'm pretty sure I was using 1024 byte messages for these.tests. > > I think it would be useful if I made the source available when I get back to the UK at the end of the week. > Yes, thanks. I'll be on vacation from end of this week till new year so it may not get resolved till then, but I definitely want to get to the bottom of this. Cheers, Alan. From aconway at redhat.com Tue Dec 11 13:39:02 2007 From: aconway at redhat.com (Alan Conway) Date: Tue, 11 Dec 2007 08:39:02 -0500 Subject: [Rhm-users] RHM performance In-Reply-To: <475D94B2.3090205@redhat.com> References: <001a01c83983$ecbb8370$c6328a50$@com> <475D94B2.3090205@redhat.com> Message-ID: <475E92F6.1060204@redhat.com> Gordon Sim wrote: > Andy Grove wrote: >> The client that starts with a single thread and then adds a new thread >> every 5 seconds. Each thread continuously publishes a fixed size >> message to one of 50 topics at random. Throughput is measured on the >> client for each 5 second duration (the cost of setting up each thread >> and connecting to JMS is excluded from the throughput measurement). > > I suspect a large part of the difference between the measurements for > the two systems may be attributable to the fact that publish is > synchronous for durable messages in 0-10 (the next message is not sent > until the broker acknowledges the previous one) whereas in 0-8 there is > no confirmation of the publication (the fact that send() returns does > not necessarily imply that the broker has even received it, let alone > stored it). > From my reading of Andy's mail, I thought he was comparing the Qpid trunk Java broker with RHM (essentially the Qpid trunk C++ broker) using the same Java client for both. In that case both tests are using 0-10. Andy can you confirm or deny my understanding? I also suspect that different ack behavior or something of that sort is likely to be the problem - something that got implemented/interpreted differently in the two brokers and shows up in this test case particularly for some reason. Cheers, Alan. From andy at codesuccess.com Tue Dec 11 13:46:37 2007 From: andy at codesuccess.com (=?utf-8?B?QW5keSBHcm92ZQ==?=) Date: Tue, 11 Dec 2007 13:46:37 +0000 Subject: [Rhm-users] RHM performance In-Reply-To: <475E92F6.1060204@redhat.com> References: <001a01c83983$ecbb8370$c6328a50$@com> <475D94B2.3090205@redhat.com><475E92F6.1060204@redhat.com> Message-ID: <1894183504-1197380812-cardhu_decombobulator_blackberry.rim.net-627259921-@bxe013.bisx.produk.on.blackberry> I was comparing qpid java broker from the M2 release with RHM C++ broker from the MRG beta. Thanks. Sent from my BlackBerry? wireless device -----Original Message----- From: Alan Conway Date: Tue, 11 Dec 2007 08:39:02 To:Gordon Sim Cc:Andy Grove , rhm-users at redhat.com Subject: Re: [Rhm-users] RHM performance Gordon Sim wrote: > Andy Grove wrote: >> The client that starts with a single thread and then adds a new thread >> every 5 seconds. Each thread continuously publishes a fixed size >> message to one of 50 topics at random. Throughput is measured on the >> client for each 5 second duration (the cost of setting up each thread >> and connecting to JMS is excluded from the throughput measurement). > > I suspect a large part of the difference between the measurements for > the two systems may be attributable to the fact that publish is > synchronous for durable messages in 0-10 (the next message is not sent > until the broker acknowledges the previous one) whereas in 0-8 there is > no confirmation of the publication (the fact that send() returns does > not necessarily imply that the broker has even received it, let alone > stored it). > From my reading of Andy's mail, I thought he was comparing the Qpid trunk Java broker with RHM (essentially the Qpid trunk C++ broker) using the same Java client for both. In that case both tests are using 0-10. Andy can you confirm or deny my understanding? I also suspect that different ack behavior or something of that sort is likely to be the problem - something that got implemented/interpreted differently in the two brokers and shows up in this test case particularly for some reason. Cheers, Alan. From gsim at redhat.com Tue Dec 11 13:49:49 2007 From: gsim at redhat.com (Gordon Sim) Date: Tue, 11 Dec 2007 13:49:49 +0000 Subject: [Rhm-users] RHM performance In-Reply-To: <475E92F6.1060204@redhat.com> References: <001a01c83983$ecbb8370$c6328a50$@com> <475D94B2.3090205@redhat.com> <475E92F6.1060204@redhat.com> Message-ID: <475E957D.3040206@redhat.com> Alan Conway wrote: > From my reading of Andy's mail, I thought he was comparing the Qpid > trunk Java broker with RHM (essentially the Qpid trunk C++ broker) using > the same Java client for both. In that case both tests are using 0-10. There is no qpid java broker that supports 0-10, the java broker on trunk speaks 0-8. The java client on trunk can speak both 0-8 and 0-10. When communicating with the java broker it will use 0-8, when communicating with the c++ broker it will communicate using 0-10. From andy at codesuccess.com Tue Dec 11 13:49:27 2007 From: andy at codesuccess.com (=?utf-8?B?QW5keSBHcm92ZQ==?=) Date: Tue, 11 Dec 2007 13:49:27 +0000 Subject: [Rhm-users] RHM performance In-Reply-To: <475E957D.3040206@redhat.com> References: <001a01c83983$ecbb8370$c6328a50$@com> <475D94B2.3090205@redhat.com> <475E92F6.1060204@redhat.com><475E957D.3040206@redhat.com> Message-ID: <621664320-1197380982-cardhu_decombobulator_blackberry.rim.net-551722470-@bxe013.bisx.produk.on.blackberry> Just to clarify my usage - the client was using the jars from the rhm beta release in both cases. Thanks. Sent from my BlackBerry? wireless device -----Original Message----- From: Gordon Sim Date: Tue, 11 Dec 2007 13:49:49 To:Alan Conway Cc:Andy Grove , rhm-users at redhat.com Subject: Re: [Rhm-users] RHM performance Alan Conway wrote: > From my reading of Andy's mail, I thought he was comparing the Qpid > trunk Java broker with RHM (essentially the Qpid trunk C++ broker) using > the same Java client for both. In that case both tests are using 0-10. There is no qpid java broker that supports 0-10, the java broker on trunk speaks 0-8. The java client on trunk can speak both 0-8 and 0-10. When communicating with the java broker it will use 0-8, when communicating with the c++ broker it will communicate using 0-10. From cctrieloff at redhat.com Tue Dec 11 13:54:23 2007 From: cctrieloff at redhat.com (Carl Trieloff) Date: Tue, 11 Dec 2007 08:54:23 -0500 Subject: [Rhm-users] RHM performance In-Reply-To: <1894183504-1197380812-cardhu_decombobulator_blackberry.rim.net-627259921-@bxe013.bisx.produk.on.blackberry> References: <001a01c83983$ecbb8370$c6328a50$@com> <475D94B2.3090205@redhat.com><475E92F6.1060204@redhat.com> <1894183504-1197380812-cardhu_decombobulator_blackberry.rim.net-627259921-@bxe013.bisx.produk.on.blackberry> Message-ID: <475E968F.4010302@redhat.com> I think this is what Gordon pointed out - async default mode on M2 java client, sync default on trunk java client. we will confirm and let you know. Carl. Andy Grove wrote: > I was comparing qpid java broker from the M2 release with RHM C++ broker from the MRG beta. > > Thanks. > > Sent from my BlackBerry? wireless device > > -----Original Message----- > From: Alan Conway > > Date: Tue, 11 Dec 2007 08:39:02 > To:Gordon Sim > Cc:Andy Grove , rhm-users at redhat.com > Subject: Re: [Rhm-users] RHM performance > > > Gordon Sim wrote: > >> Andy Grove wrote: >> >>> The client that starts with a single thread and then adds a new thread >>> every 5 seconds. Each thread continuously publishes a fixed size >>> message to one of 50 topics at random. Throughput is measured on the >>> client for each 5 second duration (the cost of setting up each thread >>> and connecting to JMS is excluded from the throughput measurement). >>> >> I suspect a large part of the difference between the measurements for >> the two systems may be attributable to the fact that publish is >> synchronous for durable messages in 0-10 (the next message is not sent >> until the broker acknowledges the previous one) whereas in 0-8 there is >> no confirmation of the publication (the fact that send() returns does >> not necessarily imply that the broker has even received it, let alone >> stored it). >> >> > > From my reading of Andy's mail, I thought he was comparing the Qpid > trunk Java broker with RHM (essentially the Qpid trunk C++ broker) using > the same Java client for both. In that case both tests are using 0-10. > > Andy can you confirm or deny my understanding? > > I also suspect that different ack behavior or something of that sort is > likely to be the problem - something that got implemented/interpreted > differently in the two brokers and shows up in this test case > particularly for some reason. > > Cheers, > Alan. > > _______________________________________________ > Rhm-users mailing list > Rhm-users at redhat.com > https://www.redhat.com/mailman/listinfo/rhm-users > -------------- next part -------------- An HTML attachment was scrubbed... URL: From aconway at redhat.com Tue Dec 11 14:03:36 2007 From: aconway at redhat.com (Alan Conway) Date: Tue, 11 Dec 2007 09:03:36 -0500 Subject: [Rhm-users] RHM performance In-Reply-To: <475E957D.3040206@redhat.com> References: <001a01c83983$ecbb8370$c6328a50$@com> <475D94B2.3090205@redhat.com> <475E92F6.1060204@redhat.com> <475E957D.3040206@redhat.com> Message-ID: <475E98B8.5030404@redhat.com> Gordon Sim wrote: > Alan Conway wrote: >> From my reading of Andy's mail, I thought he was comparing the Qpid >> trunk Java broker with RHM (essentially the Qpid trunk C++ broker) >> using the same Java client for both. In that case both tests are using >> 0-10. > > There is no qpid java broker that supports 0-10, the java broker on > trunk speaks 0-8. The java client on trunk can speak both 0-8 and 0-10. > When communicating with the java broker it will use 0-8, when > communicating with the c++ broker it will communicate using 0-10. > Ah, good, I'm talking nonsense. That probably makes it easier to track down the problem :)