Re: [Firebird-devel] True SMP support in SuperClassic and V3
Fra: Alex Peshkoff [mailto:peshk...@mail.ru] On 11/13/12 18:18, Poul Dige wrote: I could put a screen dump here of the Windows 2008R2/63 task manager with AMD Opteron 6274, 2x16 core running FB2.5.1 SC/32 bit (due to some 32 bit UDF). I don't know if you are interested at a gaze. We see exactly the same kind of usage, 8 cores are in use and 24 are more or less doing nothing. However, the cores in use are not maxed out - so it COULD be something with power management that the OS doesn't want to activate more CPU's than necessary. I can't tell about that for sure. Is it SC or CS? Hi Alex, it is Super Classic. Let me pay your attention that in initially described case use of classic made all cores loaded. How does your server behave with classic? Hi Alex, It is a very important production server and I can't find a good way to make tests on it by switching to classic. Thing is, the connections are very short lived, so it is a nice feature that the threads are spawn very quickly by Firebird. I have seen another, similar server (32 cores Win2k8R2) with FB2.1 CS which seems to distribute load evenly, so I guess the problem lies in thread handling - not if FB but in Windows itself. I will try to install an update for Windows as it seems to be a well known problem with Windows 7/2008R2 vs AMD Bulldozer architecture, so the following hotfix may just be a solution (there are two hotfixes, the other one is referred to in the article): http://support.microsoft.com/kb/2646060 I will re-post once we have had opportunity to update the server, which will probably not be until the weekend. I installed the above mentioned hotfix from MS (toghether with the prerequisit KB2645594) this past weekend. Today I see: - exactly the same. 55 threads (according to task manager) in FB 2.5.1.26351 SC (Super Classic) spread over only 8 of the 32 cores on the Win2008R2 SP1 server (see hw specs above). The cores aren't max'ed out, so at least in princip there is an excuse for not firing up the remaining 24 cores. But, somehow I'd be more happy to see an equal load on all cores. In other words - the hotfix didn't change anything noticably. Does FB decide anything regarding on which cores to run, or is it entirely a Windows decision? I imagine the latter, but of course we have the affinity parameter in the config file for super server installations, so at least it is possible to influence choice of processor core. Are there any suggestions what I should try to do next? We have actually 2 similar servers, both running 2.5.1SC (for now). The other server COULD be equipped with CS instead, if that would make a lot of sence as a trial. Even SS, as it primarily serves a lot of different databases with only a few connections on each. Your call, Alex :) Best regards Poul -- Monitor your physical, virtual and cloud infrastructure from a single web console. Get in-depth insight into apps, servers, databases, vmware, SAP, cloud infrastructure, etc. Download 30-day Free Trial. Pricing starts from $795 for 25 servers or applications! http://p.sf.net/sfu/zoho_dev2dev_nov Firebird-Devel mailing list, web interface at https://lists.sourceforge.net/lists/listinfo/firebird-devel
Re: [Firebird-devel] True SMP support in SuperClassic and V3
On 11/19/12 18:59, Poul Dige wrote: Fra: Alex Peshkoff [mailto:peshk...@mail.ru] On 11/13/12 18:18, Poul Dige wrote: I could put a screen dump here of the Windows 2008R2/63 task manager with AMD Opteron 6274, 2x16 core running FB2.5.1 SC/32 bit (due to some 32 bit UDF). I don't know if you are interested at a gaze. We see exactly the same kind of usage, 8 cores are in use and 24 are more or less doing nothing. However, the cores in use are not maxed out - so it COULD be something with power management that the OS doesn't want to activate more CPU's than necessary. I can't tell about that for sure. Is it SC or CS? Hi Alex, it is Super Classic. Let me pay your attention that in initially described case use of classic made all cores loaded. How does your server behave with classic? Hi Alex, It is a very important production server and I can't find a good way to make tests on it by switching to classic. Thing is, the connections are very short lived, so it is a nice feature that the threads are spawn very quickly by Firebird. I have seen another, similar server (32 cores Win2k8R2) with FB2.1 CS which seems to distribute load evenly, so I guess the problem lies in thread handling - not if FB but in Windows itself. I will try to install an update for Windows as it seems to be a well known problem with Windows 7/2008R2 vs AMD Bulldozer architecture, so the following hotfix may just be a solution (there are two hotfixes, the other one is referred to in the article): http://support.microsoft.com/kb/2646060 I will re-post once we have had opportunity to update the server, which will probably not be until the weekend. I installed the above mentioned hotfix from MS (toghether with the prerequisit KB2645594) this past weekend. Today I see: - exactly the same. 55 threads (according to task manager) in FB 2.5.1.26351 SC (Super Classic) spread over only 8 of the 32 cores on the Win2008R2 SP1 server (see hw specs above). The cores aren't max'ed out, so at least in princip there is an excuse for not firing up the remaining 24 cores. But, somehow I'd be more happy to see an equal load on all cores. Ahh - they are not full loaded? Probably I did not notice that fact before, sorry. In that case I suppose windows does the best for your performance - keeping less cores active lets them run at higher frequency (if I understand correctly how does your CPU work). In other words - the hotfix didn't change anything noticably. Does FB decide anything regarding on which cores to run, or is it entirely a Windows decision? I imagine the latter, but of course we have the affinity parameter in the config file for super server installations, so at least it is possible to influence choice of processor core. Are there any suggestions what I should try to do next? We have actually 2 similar servers, both running 2.5.1SC (for now). The other server COULD be equipped with CS instead, if that would make a lot of sence as a trial. Even SS, as it primarily serves a lot of different databases with only a few connections on each. Your call, Alex :) If 8 cores are not 100% loaded, this becomes more windows + amd rather than firebird issue. I give up. -- Monitor your physical, virtual and cloud infrastructure from a single web console. Get in-depth insight into apps, servers, databases, vmware, SAP, cloud infrastructure, etc. Download 30-day Free Trial. Pricing starts from $795 for 25 servers or applications! http://p.sf.net/sfu/zoho_dev2dev_nov Firebird-Devel mailing list, web interface at https://lists.sourceforge.net/lists/listinfo/firebird-devel
Re: [Firebird-devel] True SMP support in SuperClassic and V3
On 19-11-2012 14:39, Alex Peshkoff wrote: Are there any suggestions what I should try to do next? We have actually 2 similar servers, both running 2.5.1SC (for now). The other server COULD be equipped with CS instead, if that would make a lot of sence as a trial. Even SS, as it primarily serves a lot of different databases with only a few connections on each. Your call, Alex :) If 8 cores are not 100% loaded, this becomes more windows + amd rather than firebird issue. I give up. What was being tested? There may be no sense to load all cores and let them wait for disk IO. Or was the benchmark done with something pure CPU-bound (a stored procedure doing math in a loop, for example)? Adriano -- Monitor your physical, virtual and cloud infrastructure from a single web console. Get in-depth insight into apps, servers, databases, vmware, SAP, cloud infrastructure, etc. Download 30-day Free Trial. Pricing starts from $795 for 25 servers or applications! http://p.sf.net/sfu/zoho_dev2dev_nov Firebird-Devel mailing list, web interface at https://lists.sourceforge.net/lists/listinfo/firebird-devel
Re: [Firebird-devel] True SMP support in SuperClassic and V3
-Oprindelig meddelelse- Fra: Adriano dos Santos Fernandes [mailto:adrian...@gmail.com] Sendt: 19. november 2012 17:47 Til: For discussion among Firebird Developers Emne: Re: [Firebird-devel] True SMP support in SuperClassic and V3 On 19-11-2012 14:39, Alex Peshkoff wrote: Are there any suggestions what I should try to do next? We have actually 2 similar servers, both running 2.5.1SC (for now). The other server COULD be equipped with CS instead, if that would make a lot of sence as a trial. Even SS, as it primarily serves a lot of different databases with only a few connections on each. Your call, Alex :) If 8 cores are not 100% loaded, this becomes more windows + amd rather than firebird issue. I give up. What was being tested? There may be no sense to load all cores and let them wait for disk IO. Or was the benchmark done with something pure CPU-bound (a stored procedure doing math in a loop, for example)? Adriano The test is quite CPU bound, the disk I/O was relatively low (and running on a FusionIO SSD). I do think that this is an AMD/Windows issue, too. It was just for testing. And Doychin is right about the NUMA-architecture, so the (total of) 32 cores are distributed into 4 NUMA nodes. That could explain the behaviour that one NUMA node is preferred to minimize memory sharing. The dual processor Opteron has two memory banks, which is again optimally equipped with an equal number of memory modules - maybe to satisfy each NUMA node with its own memory bank. I wonder, though, if this is optimal for, say, super server with many different DB's. In that case the optimal solution - I think - is to have one DB per core, if possible, and not try to keep the threads together on one NUMA node. But still that is not a FB problem. Kind regards Poul -- Monitor your physical, virtual and cloud infrastructure from a single web console. Get in-depth insight into apps, servers, databases, vmware, SAP, cloud infrastructure, etc. Download 30-day Free Trial. Pricing starts from $795 for 25 servers or applications! http://p.sf.net/sfu/zoho_dev2dev_nov Firebird-Devel mailing list, web interface at https://lists.sourceforge.net/lists/listinfo/firebird-devel
Re: [Firebird-devel] True SMP support in SuperClassic and V3
Fra: Dmitry Yemanov [mailto:firebi...@yandex.ru] 12.11.2012 16:52, vince.dug...@virginactive.co.za wrote: When running under load (150+ busy connections) SuperClassic clearly uses only one NUMA node (threads are spread across 16 cores), while Classic is spread evenly across all 64 cores. Weird, I'd expect them behaving the same way. Perhaps this is somewhat OS related. How is this going to run in V3? Will we see full use of all cores on all processors, or will the behaviour be as in SuperClassic. Supposedly, it would be the same as in SuperClassic. But so far it doesn't look like a Firebird issue, it belongs elsewhere. Is this a machine architectural issue? i.e. the NUMA node architecture makes it a little more difficult for one NUMA node to access the memory 'allocated' to another NUMA node? A little more difficult - true, but it should mean a little slower, not preventing that completely. I could put a screen dump here of the Windows 2008R2/63 task manager with AMD Opteron 6274, 2x16 core running FB2.5.1 SC/32 bit (due to some 32 bit UDF). I don't know if you are interested at a gaze. We see exactly the same kind of usage, 8 cores are in use and 24 are more or less doing nothing. However, the cores in use are not maxed out - so it COULD be something with power management that the OS doesn't want to activate more CPU's than necessary. I can't tell about that for sure. Best regards Poul -- Monitor your physical, virtual and cloud infrastructure from a single web console. Get in-depth insight into apps, servers, databases, vmware, SAP, cloud infrastructure, etc. Download 30-day Free Trial. Pricing starts from $795 for 25 servers or applications! http://p.sf.net/sfu/zoho_dev2dev_nov Firebird-Devel mailing list, web interface at https://lists.sourceforge.net/lists/listinfo/firebird-devel
Re: [Firebird-devel] True SMP support in SuperClassic and V3
On Tue, 13 Nov 2012 14:18:22 +, Poul Dige p...@tabulex.dk wrote: I could put a screen dump here of the Windows 2008R2/63 task manager with AMD Opteron 6274, 2x16 core running FB2.5.1 SC/32 bit (due to some 32 bit UDF). I don't know if you are interested at a gaze. We see exactly the same kind of usage, 8 cores are in use and 24 are more or less doing nothing. However, the cores in use are not maxed out - so it COULD be something with power management that the OS doesn't want to activate more CPU's than necessary. I can't tell about that for sure. I know that Windows 8 will 'park' cores if there is no work for them, and it will try to assign threads to the active/unparked cores first. When the load exceeds a certain threshold, then it will unpark an additional core and schedule threads to that (at least that is what I observed on my Windows 8 Pro on an 8 'core'(*) FX-8120 at home), as far as I know Windows 2008 server also has similar scheduling algorithms (depending on the processor). The reason is simple: it conserves some power, and some processor have a kind of speed boost when not all cores are loaded (ie: it will 'overclock' the active cores because there is more thermal 'headroom' as not all cores are powered and generating heat). This allows for faster processing if work isn't really multithreaded, and doesn't actually benefit much from running multicore. (*): Actually the AMD FX is not fully 8 core, it is more like a 4 core which allows 2 threads to be scheduled simultaneously by having same features/instructions twice, and other features/instructions only once per core. Mark -- Monitor your physical, virtual and cloud infrastructure from a single web console. Get in-depth insight into apps, servers, databases, vmware, SAP, cloud infrastructure, etc. Download 30-day Free Trial. Pricing starts from $795 for 25 servers or applications! http://p.sf.net/sfu/zoho_dev2dev_nov Firebird-Devel mailing list, web interface at https://lists.sourceforge.net/lists/listinfo/firebird-devel
Re: [Firebird-devel] True SMP support in SuperClassic and V3
Fra: Mark Rotteveel [mailto:m...@lawinegevaar.nl] Emne: Re: [Firebird-devel] True SMP support in SuperClassic and V3 On Tue, 13 Nov 2012 14:18:22 +, Poul Dige p...@tabulex.dk wrote: I could put a screen dump here of the Windows 2008R2/63 task manager with AMD Opteron 6274, 2x16 core running FB2.5.1 SC/32 bit (due to some 32 bit UDF). I don't know if you are interested at a gaze. We see exactly the same kind of usage, 8 cores are in use and 24 are more or less doing nothing. However, the cores in use are not maxed out - so it COULD be something with power management that the OS doesn't want to activate more CPU's than necessary. I can't tell about that for sure. I know that Windows 8 will 'park' cores if there is no work for them, and it will try to assign threads to the active/unparked cores first. When the load exceeds a certain threshold, then it will unpark an additional core and schedule threads to that (at least that is what I observed on my Windows 8 Pro on an 8 'core'(*) FX-8120 at home), as far as I know Windows 2008 server also has similar scheduling algorithms (depending on the processor). The reason is simple: it conserves some power, and some processor have a kind of speed boost when not all cores are loaded (ie: it will 'overclock' the active cores because there is more thermal 'headroom' as not all cores are powered and generating heat). This allows for faster processing if work isn't really multithreaded, and doesn't actually benefit much from running multicore. (*): Actually the AMD FX is not fully 8 core, it is more like a 4 core which allows 2 threads to be scheduled simultaneously by having same features/instructions twice, and other features/instructions only once per core. Mark Maybe this article about core parking will be of interest - amazing, what you can Google nowadays: http://support.microsoft.com/kb/2646060 Regards, Poul -- Monitor your physical, virtual and cloud infrastructure from a single web console. Get in-depth insight into apps, servers, databases, vmware, SAP, cloud infrastructure, etc. Download 30-day Free Trial. Pricing starts from $795 for 25 servers or applications! http://p.sf.net/sfu/zoho_dev2dev_nov Firebird-Devel mailing list, web interface at https://lists.sourceforge.net/lists/listinfo/firebird-devel
[Firebird-devel] True SMP support in SuperClassic and V3
Hi all, I have been doing a lot of testing on a monster Windows server so that we can move from a Solaris 2.1.3 build (thanks Paul) to 2.5 on preferably Windows, or perhaps Linux. The box has 8 processors, each with 8 cores. These are arranged into 4 NUMA nodes, each with 2 processors. Windows reports 4 processors (i.e. each NUMA node is seen as one processor) , each with 16 cores. 128GB RAM, Windows 2008 R2. Firebird Version: WI-V2.5.2.26539 Firebird 2.5 My concern is that in this article: http://www.firebirdsql.org/manual/qsg25-classic-or-super.html it is stated that SuperServer uses one processor (unless CPUAffinityMask is changed) and that Classic and SuperClassic ignore CPUAffinityMask, and use all processors. When running under load (150+ busy connections) SuperClassic clearly uses only one NUMA node (threads are spread across 16 cores), while Classic is spread evenly across all 64 cores. CPUAffinityMask makes no difference (as stated in the docs). How is this going to run in V3? Will we see full use of all cores on all processors, or will the behaviour be as in SuperClassic. Is this a machine architectural issue? i.e. the NUMA node architecture makes it a little more difficult for one NUMA node to access the memory 'allocated' to another NUMA node? which implies that V3's shared cache may be restricted to one node, and not spread across all nodes? Regards Vince Vince Duggan I.T. Architect Virgin Active South Africa (Pty) Ltd Tel (+27) (0)21 684 3525 Fax (+27) (0)21 684 3225 Cell (+27) (0)82 747 6127 Email: vince.dug...@virginactive.co.za Live happily ever active -- This message has been scanned for viruses and dangerous content by MailScanner, and is believed to be clean. -- Everyone hates slow websites. So do we. Make your web apps faster with AppDynamics Download AppDynamics Lite for free today: http://p.sf.net/sfu/appdyn_d2d_novFirebird-Devel mailing list, web interface at https://lists.sourceforge.net/lists/listinfo/firebird-devel
Re: [Firebird-devel] True SMP support in SuperClassic and V3
I might have a theory on that subject. With HT you share the same execution blocks among two pipelines (meaning you are trying to balance twice as many threads/instructions and so on, without increasing reorder caches/registers/data and instruction caches and other important factors). On a normal PC you have threads with different profiles - media streaming, integer calculations and so on, so most tasks don`t overlap as much and you feel a big boost. With the database (or any other dedicated server for that matter), on the other hand, you have basically the same profile for all threads and you starve the CPU from certain blocks/registers, while you don`t really utilize the others. You also have some kind of synchronization between the two pipelines. When you have a single user (filling the info in the database) you don`t feel the difference because it is basically the same CPU core doing the work. I could be wrong on that one, but under ultra heavy loads I guess all those things will be noticeable. Nobody was able to explain why such big difference. Interesting that the performance was similar in both machines when loading the database with information. The huge difference only appeared when using the database. []s Carlos http://www.firebirdnews.org FireBase - http://www.FireBase.com.br -- Using Opera's revolutionary email client: http://www.opera.com/mail/ -- Everyone hates slow websites. So do we. Make your web apps faster with AppDynamics Download AppDynamics Lite for free today: http://p.sf.net/sfu/appdyn_d2d_nov Firebird-Devel mailing list, web interface at https://lists.sourceforge.net/lists/listinfo/firebird-devel
Re: [Firebird-devel] True SMP support in SuperClassic and V3
12.11.2012 16:52, vince.dug...@virginactive.co.za wrote: When running under load (150+ busy connections) SuperClassic clearly uses only one NUMA node (threads are spread across 16 cores), while Classic is spread evenly across all 64 cores. Weird, I'd expect them behaving the same way. Perhaps this is somewhat OS related. How is this going to run in V3? Will we see full use of all cores on all processors, or will the behaviour be as in SuperClassic. Supposedly, it would be the same as in SuperClassic. But so far it doesn't look like a Firebird issue, it belongs elsewhere. Is this a machine architectural issue? i.e. the NUMA node architecture makes it a little more difficult for one NUMA node to access the memory 'allocated' to another NUMA node? A little more difficult - true, but it should mean a little slower, not preventing that completely. Dmitry -- Everyone hates slow websites. So do we. Make your web apps faster with AppDynamics Download AppDynamics Lite for free today: http://p.sf.net/sfu/appdyn_d2d_nov Firebird-Devel mailing list, web interface at https://lists.sourceforge.net/lists/listinfo/firebird-devel
Re: [Firebird-devel] True SMP support in SuperClassic and V3
Hi all, Thanks for the replies and comments. Vince, FB 3 snapshot is already available for download, so you can check it by yourself (and if so, please report the results here). I have V3 running on a much smaller VM, just to test if my stress tests run at all - short answer : they don't :-) The DB is a Dialect 1 with a lot of legacy rubbish. I will certainly update the tests in order to get them to run properly on V3, and will report back. The big box is earmarked as a production server, so I won't have access for long, but I will make a plan. I should be able to run both 2.5 and 3 on the same box? Vince Actually, with such heavy server, I think you can help the Project a lot with FB 3 tests. Personally, I got weird results using FB 3 in my notebook (4 cores, 2 real and 2 HT and 8GB RAM) and TPC-C tests. The same tests showed much better results (at last 10x better) in a quad core desktop (this time, 4 real cores but only 4GB RAM). Nobody was able to explain why such big difference. Interesting that the performance was similar in both machines when loading the database with information. The huge difference only appeared when using the database. -- This message has been scanned for viruses and dangerous content by MailScanner, and is believed to be clean. -- Everyone hates slow websites. So do we. Make your web apps faster with AppDynamics Download AppDynamics Lite for free today: http://p.sf.net/sfu/appdyn_d2d_novFirebird-Devel mailing list, web interface at https://lists.sourceforge.net/lists/listinfo/firebird-devel
Re: [Firebird-devel] True SMP support in SuperClassic and V3
On 11/12/12 19:18, vince.dug...@virginactive.co.za wrote: Hi all, Thanks for the replies and comments. Vince, FB 3 snapshot is already available for download, so you can check it by yourself (and if so, please report the results here). I have V3 running on a much smaller VM, just to test if my stress tests run at all - short answer : they don't :-) The DB is a Dialect 1 with a lot of legacy rubbish. I will certainly update the tests in order to get them to run properly on V3, and will report back. The big box is earmarked as a production server, so I won't have access for long, but I will make a plan. I should be able to run both 2.5 and 3 on the same box? Yes, certainly - but you need some manual configuration to be done and on my mind it will be very good idea to try this on non-production box first. -- Everyone hates slow websites. So do we. Make your web apps faster with AppDynamics Download AppDynamics Lite for free today: http://p.sf.net/sfu/appdyn_d2d_nov Firebird-Devel mailing list, web interface at https://lists.sourceforge.net/lists/listinfo/firebird-devel
Re: [Firebird-devel] True SMP support in SuperClassic and V3
IA When you have a single user (filling the info in the database) you don`t IA feel the difference because it is basically the same CPU core doing the IA work. Here is more information (from my logs and memory): The LOAD was set to simulate 10 users inserting information at the same time so, there was 10 connections and 10 threads acting at the same time to fill up the DB. Some weird results also appeared when comparing FB 3 TPC-C results in the notebook. FB 3 performance was 50% worse compared to FB 2.5 SS/CS/SC. From my logs: 2.5.1 CS - TPC-C Throughput: 1390.19 tpmC 2.5.1 SC - TPC-C Throughput: 1221.92 tpmC - Affinity set to 1 2.5.1 SS - TPC-C Throughput: 1306.06 tpmC 2.5.2 CS - TPC-C Throughput: 1144.95 tpmC 2.5.2 SC - TPC-C Throughput: 1156.87 tpmC - Affinity set to 1 2.5.2 SS - TPC-C Throughput: 1358.34 tpmC 3.0 SS - TPC-C Throughput: 514.43 tpmC (SharedCache was true and SharesDatabase was false) Notebook config: Windows 7 Pro 64bits Intel Core i5 2410M 500GB Sata 7200rpm 8GB RAM I agree that in a real scenario, a notebook would not be used as a server, but I wonder why such bad performance, even when compared to FB 2.5.2 SS attached to single CPU. Curious that the LOAD process seems to not suffer from the problem. I just hope nothing similar will show up in real servers, since it would be very disappointing to the user if he moves to FB 3 expecting huge improvement and find out that it started to behave worse than FB 2.5. I plan to start a global testing campaign at FirebirdNews.org when FB 3 alpha shows up, in the hope to detect all possible bad behaviors and bugs, in the hope to speed up the release process. []s Carlos http://www.firebirdnews.org FireBase - http://www.FireBase.com.br -- Everyone hates slow websites. So do we. Make your web apps faster with AppDynamics Download AppDynamics Lite for free today: http://p.sf.net/sfu/appdyn_d2d_nov Firebird-Devel mailing list, web interface at https://lists.sourceforge.net/lists/listinfo/firebird-devel
Re: [Firebird-devel] True SMP support in SuperClassic and V3
Sorry, the affinity setting was about SS, not SC. Wrong copy/paste ;-) []s Carlos http://www.firebirdnews.org FireBase - http://www.FireBase.com.br IA When you have a single user (filling the info in the database) you don`t IA feel the difference because it is basically the same CPU core doing the IA work. CHC Here is more information (from my logs and memory): CHC The LOAD was set to simulate 10 users inserting information at the same CHC time so, there was 10 connections and 10 threads acting at the same CHC time to fill up the DB. CHC Some weird results also appeared when comparing FB 3 TPC-C results in CHC the notebook. FB 3 performance was 50% worse compared to FB 2.5 CHC SS/CS/SC. From my logs: CHC 2.5.1 CS - TPC-C Throughput: 1390.19 tpmC CHC 2.5.1 SC - TPC-C Throughput: 1221.92 tpmC - Affinity set to 1 CHC 2.5.1 SS - TPC-C Throughput: 1306.06 tpmC CHC 2.5.2 CS - TPC-C Throughput: 1144.95 tpmC CHC 2.5.2 SC - TPC-C Throughput: 1156.87 tpmC - Affinity set to 1 CHC 2.5.2 SS - TPC-C Throughput: 1358.34 tpmC CHC 3.0 SS - TPC-C Throughput: 514.43 tpmC (SharedCache was true and SharesDatabase was false) -- Everyone hates slow websites. So do we. Make your web apps faster with AppDynamics Download AppDynamics Lite for free today: http://p.sf.net/sfu/appdyn_d2d_nov Firebird-Devel mailing list, web interface at https://lists.sourceforge.net/lists/listinfo/firebird-devel