search for: netbench

Hi All, I'm running netbench against our samba based filer and having I believe a controller problem. When I configure the test to run multiple engines per client (about 5 in my case) and about 20 clients so all together I have 100 engines, the controller crashes. My clients are a mix of NT4, winxp and win2000 systems....

I have installed the latest alpha of samba (1.9.18alpha4) with the intent of testing against the Ziff-Davis benchmark NetBench. The compile and install went OK, and I can see the server in the browser and connect to the public share I created (logins work OK). Running the tests in DIAGNOSIS.txt produces no errors also. I can also install netbench and start up the controller from the binary that resides on the share. B...

Hi all, there is a problem with the 'Default: nt smb support = yes' option in smb.conf for at least Samba 2.2.3a and 2.2.4 (on FreeBSD 4.5-STABLE) when using Netbench 7.0.2 on WXP or W2k clients. If the client is set to run 2+ Netbench engines (to simulate 2+ real clients) and 'nt smb support = yes', all but one engine fail pretty quickly, usually during a movefile oder deletefile operation (as reported by the failing engine). This does not happen with...

Has anyone gotten Netbench 6.0 running on Windows 2000 clients to work under samba? I'm running Samba 2.1.0-prealpha on linux with Windows 2000 professional clients. When I try to run it, it initializes successfully but when it starts to execute a test I get the error "Mix lock of handle 6 for length 1 at offset 2...

-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 I'm running some netbench tests, and I have some questions I'd like to ask. First, what is the difference between the dm.tst and ent_dm.tst NetBench suites? Second, what results have people been getting with Samba 3.0.0? I'm seeing a peak at 340.00 for ent_dm.tst and 101.00 for dm.tst on a P4 system with 512 MB RA...

-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 I was running a netbench performance test against Samba 3.0.11, and the test has not been able to complete fully. The samba logs contain lines such as this: [2005/02/28 11:35:23, 1] smbd/service.c:make_connection_snum(645) ~ slave16 (192.168.133.116) connect to service netbench initially as user BENCHLAB+slave16 (uid=10...

...g gisel-with-localizer O0 -g gisel-without-localizer > SingleSource/Benchmarks/Misc/perlin 14272 14640 18344 25.95% > SingleSource/Benchmarks/Dhrystone/dry 16560 17144 20160 18.21% > SingleSource/Benchmarks/Stanford/QueensProfile 13912 14192 15136 6.79% > MultiSource/Benchmarks/Trimaran/netbench-url/netbench-url 71400 72272 75504 4.53% > > I haven't had time to investigate what exact changes make the code size go up that much with the localizer pass in those cases... > >> >> The only thing I can think of is that we duplicate constants that are expensive to mater...

...g gisel-with-localizer O0 -g gisel-without-localizer > SingleSource/Benchmarks/Misc/perlin 14272 14640 18344 25.95% > SingleSource/Benchmarks/Dhrystone/dry 16560 17144 20160 18.21% > SingleSource/Benchmarks/Stanford/QueensProfile 13912 14192 15136 6.79% > MultiSource/Benchmarks/Trimaran/netbench-url/netbench-url 71400 72272 > 75504 4.53% > > I haven't had time to investigate what exact changes make the code size go > up that much with the localizer pass in those cases... > > > The only thing I can think of is that we duplicate constants that are > expensive to m...

...e proposed pass-ordering is basically toward good change. > > Interesting enough, if I combine the populatePreIPOPassMgr() as the preIPO phase > (see the patch) with original populateLTOPassManager() for both IPO and postIPO, > I see significant improve to "Benchmarks/Trimaran/netbench-crc/netbench-crc" > (about 94%, 0.5665s(was) vs 0.0295s), as of I write this mail, I have not yet got chance > to figure out why this combination improves this benchmark this much. > > In teams of compile-time, the result reports my change improve the compile > time by about...

I Want to have a copmparsion with NFS v3 that shows some statistics with both attribute intensive and data intensive clients. /P -------------- next part -------------- A non-text attachment was scrubbed... Name: vcard.vcf Type: text/x-vcard Size: 392 bytes Desc: Card for Patrik Linder Url : http://lists.samba.org/archive/samba/attachments/19990301/59f94bcf/vcard.vcf

...cted files that match a > wildcarded list, similar to the wildcarded list used in the veto > files parameter. > > You might want to do this on files that you know will be heavily > contended for by clients. A good example of this is in the NetBench > SMB benchmark program, which causes heavy client contention for > files ending in .SEM. To cause Samba not to grant oplocks on these > files you would use the line (either in the [global] section or in > the section for the particular NetBen...

...rom the result, the proposed pass-ordering is basically toward good change. Interesting enough, if I combine the populatePreIPOPassMgr() as the preIPO phase (see the patch) with original populateLTOPassManager() for both IPO and postIPO, I see significant improve to "Benchmarks/Trimaran/netbench-crc/netbench-crc" (about 94%, 0.5665s(was) vs 0.0295s), as of I write this mail, I have not yet got chance to figure out why this combination improves this benchmark this much. In teams of compile-time, the result reports my change improve the compile time by about 2x, which is non-sense....

...rformance.  With ISL code generator, 22 benchmarks have >10% compile-time performance improvement over Cloog. Top 10 improvements are shown as follows: Performance Improvements - Compile Time (ISL over Cloog) MultiSource/Benchmarks/ASC_Sequoia/AMGmk/AMGmk -69.11% MultiSource/Benchmarks/Trimaran/netbench-crc/netbench-crc -44.39% SingleSource/Benchmarks/Polybench/linear-algebra/kernels/3mm/3mm -12.74% SingleSource/Benchmarks/Polybench/linear-algebra/kernels/gemm/gemm -11.21% MultiSource/Benchmarks/TSVC/LoopRerolling-flt/LoopRerolling-flt -11.14% SingleSource/Benchmarks/Polybench/linear-algebra/kerne...

...smbtorture bench.nbench from another freebsd machine. Samba 3.6 Throughput is at: 12.6165 MB/sec Samba 4.2 Throughput is at: 1.5359 MB/sec 4.2 appears almost 10X slower. What could be going wrong for me ? Is there an official bench-marking of Samba 3 and Samba 4, which I could refer to ? btw.. netbench seems to be going over SMB1 . Is there a benchmarking tool which uses SMB2/SMB3 ? Thanks, Nagendra

...uction Document. Can anyone tell me where I can find the Ziff-Davis benchmark tests? Thanks a lot, Werner Maes KULeuven "The Present Samba 2.0 was released in January 1999. One of the most significant and cool features of the 2.0 release was improved speed. Ziff-Davis Publishing used their Netbench software to benchmark Samba 2.0 on Linux against Windows NT4. They ran all of their tests on the same PC hardware, and their results showed Samba's throughput under load to be at least twice that of NT. Samba is shipped with all major Linux distributions, and Ziff-Davis tested three of those. &...

This page (of a much longer story) might be interesting : http://www.pcmag.com/article/0,2997,s%253D1474%2526a%253D16554,00.asp Netbench results show Samba 2.2 faster than Windows 2000 in all cases on the same hardware (up to 1GHz processor with 512Mb ram). Regards, Jeremy Allison, Samba Team. PS. Yes I know, lies, damn lies & benchmarks etc.... :-). PPS. I wonder if any of the Samba vendors might like to reproduce these r...

Hi all, while working with Netbench I've discovered that mangled directory names clash rather frequently with Samba 2.2.4: If I take an NT box, open the cmd prompt and make the following directories on a Samba share, client100 and client121, both mangle to CLIEN~GH. If I create directories client100 through client130 I see 11 c...

...89142,sid5_gci11 50420,00.html?track=NL-52&ad=536546) I have couple of questions regarding it: 1. Have they modified/added new features to the core CIFS, RPC, etc. protocols to support some of those features, perhaps in the DFS or other areas? 2. In terms of CIFS, they claim outstanding NetBench performance of more than 4Mbit/sec throughput. Do we know of any recent performance studies on SAMBA that compared to this. Of course, netbench measured performance depends on many things in addition to the actual SAMBA component (backend, network, clients) but I'm interested to get your th...

Andy and I briefly discussed this the other day, we have not yet got chance to list a detailed pass order for the pre- and post- IPO scalar optimizations. This is wish-list in our mind: pre-IPO: based on the ordering he propose, get rid of the inlining (or just inline tiny func), get rid of all loop xforms... post-IPO: get rid of inlining, or maybe we still need it, only

Yes, I get more than 30MB/s performance. The benchmark I use (NetBench) is essentially CPU bound, such that a faster processor = faster performance. With a very fast hardware config (dual 3.2GHz processors), I've been able to hit around 100MB/s. Changing the RAM or other attributes does not buy me much, it seems that processor power is the bottleneck (at least in...