On 10/15/08 17:01, HUGE | Rob Terhaar wrote:
> On 10/15/08 2:13 AM, "Alan M Wright" <[EMAIL PROTECTED]> wrote:
>> ...
>> As I mentioned previously, we have not enabled oplocks by default yet
>> because we haven't fully tested the implementation but we are going to
>> try to get through that testing as a high priority.
>>
>> Alan
> 
> I brought up this issue a few months ago, and Janice recommended against
> enabling oplocks on solaris cifs:
> 
> http://mail.opensolaris.org/pipermail/cifs-discuss/2008-July/000394.html

That previous recommendation seems appropriate and, I think,
consistent with my response from a few days ago:

----- Begin Included Message -----
On 10/13/08 21:41, Alan M Wright wrote:
...
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.
----- End Included Message -----

> I still think that oplocks and solaris cifs needs to be documented better.

I'm not sure how to interpret that statement.  You think that oplocks
are not documented well enough in general and neither is the Solaris
CIFS service or you are looking for more information on oplocks within
the Solaris CIFS documentation?

Oplocks are an optional part of the protocol (server's discretion) and
they should be transparent from an end-user perspective.  Performance
aside, things should work correctly without oplocks.  The caveats I
mentioned (write-ordering for journaling and the temporary delete
deadlock) arise only when oplocks are being granted, and those are
general oplock issues (not specific to an implementation) that have
been documented by Microsoft and others.

Currently, oplocks are disabled and we don't provide configuration
options to enable them.  You can enable them by going under the covers
to manipulate SMF properties and restarting the service but (just
like fiddling with registry entries on Windows) that's not a published
interface for CIFS service configuration management.

Once the oplock code has been validated, we'll need to publish and
document the options to enable/disable them.  Perhaps I'm missing your
point but I'm not sure there's anything to document until oplocks are
enabled by default in the CIFS service.

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

Reply via email to