Ulrich Windl wrote:
> (Trimmed the list of reply-to:)
> On 5 Mar 2009 at 11:42, Mike Christie wrote:
> 
> [...]
>> The second issue is that we normally do zero copy for writes. I do not 
>> think it is FS bug or net bug or a bug in the iscsi layer. Maybe more of 
>> a bug in what the user expects (who reads the man page for write() to 
>> check if the data is committed to disk when write() returns). We 
>> discussed this a couple times. For open-iscsi we tried to close the gap, 
>> by not doing zero copy writes when data digests are used. And a long 
>> long time ago this was discussed for linux-iscsi, and I think that is 
>> one of the reasons we added DID_IMM_RETRY to the scsi layer (we can then 
>> avoid the 5 retry limit in this case and retry until it is resolved).
> 
> Hi Mike,
> 
> I have no idea about the buffer handling in networking, etc, but as far as I 
> understand, you are telling the iSCSI/networking code to transmit some data 
> blocks 
> in menory via networking without making a copy of the data blocks. Now I 
> wonder if 
> there are some flags to tell the kernel not to update that data blocks, 
> because a 
> transfer is "in progress" (when actually the transfer is just queued for 
> progress). In buffer_head.h there are some flags, but I don't know what they 
> exactly mean. The other solution that comes to my mind would be a type of 
> call-
> back when a buffer was changed after being queued for transport; the call 
> back 
> would then have to re-calculate the data disgest.
> 
> Am I completely wrong here, or is it more or less the problem we are talking 
> about?

I think we are talking about the same problem.

I am not sure about some of the details in what you propose. Probably, 
the iscsi list is not the appropriate forum since we do not have a deep 
expertise in the vm or networking code. But I think if you ask on the 
other lists, they might tell you to just write your app so you do a sync 
after the write.

--~--~---------~--~----~------------~-------~--~----~
You received this message because you are subscribed to the Google Groups 
"open-iscsi" group.
To post to this group, send email to open-iscsi@googlegroups.com
To unsubscribe from this group, send email to 
open-iscsi+unsubscr...@googlegroups.com
For more options, visit this group at http://groups.google.com/group/open-iscsi
-~----------~----~----~----~------~----~------~--~---

Reply via email to