arl critique: www_veritest_com_clients_reports_microsoft_ms_netbench_pdf |
[top] [this] [up] |
date: 20030512T1420
source: http://www.veritest.com/clients/reports/microsoft/ms_netbench.pdf
|
Basically SAMBA is an application running on top of Linux. To call Red Hat Linux as Linux
is the opinion of testers.
Naturally the poit is to test Linux server suitability in Microsoft client world, but the test is quite absurd in a way, that the test does not contain Microsoft server testing as an NFS server for Linux client world - but that is not the case, but makes me wonder...
I know SAMBA .. I have been playing with it .. I do not like it at all, except for small
enviroments where files need to be served not too often, there are quite a many bad
points in samba like: But for some good words for SAMBA; it is a really amazing work of hacking, because Microsoft does not naturally assist for finding out fileserver protocol. It does not need - there are no court order to force Microsoft. Also SAMBA is normal process which means that when SAMBA goes down, only SAMBA is down, not the whole machine like one has learned in microsoftian world. The test did not contain any CPU usage graphs; Linux servers could naturally do also something else beside serving microsoftian clients. In a microsoftian world one might need to have buy an other server machine for intranet or other services, but linuxoids might want to use the fileserver as a one machine for many services. The test is done with i386 image of SAMBA .. I do not know if i686 would be any faster but still. "Test report prepared under contract from Microsoft" says much .. or does it? |
© arl