Dan,

The slow performance appears to be due to oplocks being disabled.
You can enable oplocks using svccfg but there's a caveat - we have
been testing with oplocks disabled and I have no idea whether or
not you'll encounter problems.  I'll ask for a regression cycle
with oplocks enabled but, for now, I can't provide any guarantees.

        svccfg -s smb/server
        svc:/network/smb/server> setprop smbd/oplock_enable=true
        svc:/network/smb/server> quit

You can check the setting using:

        svccfg -s smb/server listprop | grep oplock

which should produce:
        smbd/oplock_enable               boolean  true

To update the SMB service:

        svcadm restart smb/server

Alan
--

On 10/11/08 14:51, Alan M Wright wrote:
> On 10/10/08 13:08, Daniel Liebster wrote:
>> Here is a c++ program that also shows the performance gap. 
>> Completion time against a BlueArc fileserver is 9 seconds.
>> Completion time against a Windows Fileserver is 7 seconds.
>> Completion time against CIFS-server svn_98 is greater than 500 seconds
>>
>> The attached is the c source for the executable used to generate the 16MB 
>> file(genbinfile.c), and the cpp source for the executable that runs the 
>> actual test(binary.file.test.cpp).
>>
>> The screen-grab from the XP client show the first curve of activity against 
>> the Windows server, the second curve against the BlueArc, and the third long 
>> curve against the CIFS Server.
>>
>> The current version of MinGW gcc and g++ was used to compile the 
>> programs.(gcc version 3.4.5 (mingw-vista special r3))
>>
>> no flags to gcc for the file generator exe, 
>> c:\MinGW\bin\g++ -I"d:/cppPgms" -c -Wall -O binary.file.test.cpp
>> c:\MinGW\bin\g++ -obinary.file.test.cpp.exe -Wall -O
>> binary.file.test.o 
>>
>> to compile the c++ test exe.
>>
>> binary.file.test.exe then takes the location of the test file as its 
>> argument.
> 
> Dan,
> 
> Thanks.  I have this running and can see the behaviour you described.
> I'll try to find the root cause.
> 
> Alan
> _______________________________________________
> cifs-discuss mailing list
> [email protected]
> http://mail.opensolaris.org/mailman/listinfo/cifs-discuss

_______________________________________________
cifs-discuss mailing list
[email protected]
http://mail.opensolaris.org/mailman/listinfo/cifs-discuss

Reply via email to