logo       

Re: file data corruption?: msg#00226

Subject: Re: file data corruption?
Am Donnerstag, den 21.07.2005, 23:40 +0900 schrieb
hooanon05@xxxxxxxxxxx:
> thank you ming for forwarding mails.
> 
> Ming Zhang:
> > Note that the iscsi pdu in #12 is not aligned to the start of the tcp
> > segment since it contains the tail of the previous iscsi PDU starting
> > in packet #2
> 
> i have not checked if it is true, but i am afraid that splitting a
> single pdu into multiple packets is iet bug. 

As I already wrote before, iSCSI doesn't care about TCP segmentation.
Have a look at Appendix A of the RFC: "[...] The marker uses two copies
of the pointer so that a marker that spans a TCP packet boundary should
leave at least one valid copy in one of the packets. [...]"

Arne

> and i guess the root cause
> is around setting/unsetting TCP_CORK.
> should it be such like this to ensure sending a pdu?
> 
> - set CORK.
> - transmit a pdu.
> - set NDELAY.
> - unset CORK.
> - transmit data.
> 
> currently, NDELAY is not set and unsetting CORK is done after
> transmitting data.
> 
> 
> Junjiro Okajima
> 
> 
> -------------------------------------------------------
> SF.Net email is sponsored by: Discover Easy Linux Migration Strategies
> from IBM. Find simple to follow Roadmaps, straightforward articles,
> informative Webcasts and more! Get everything you need to get up to
> speed, fast. http://ads.osdn.com/?ad_id=7477&alloc_id=16492&op=click
> _______________________________________________
> Iscsitarget-devel mailing list
> Iscsitarget-devel@xxxxxxxxxxxxxxxxxxxxx
> https://lists.sourceforge.net/lists/listinfo/iscsitarget-devel
-- 
Arne Redlich
Xiranet Communications GmbH



-------------------------------------------------------
SF.Net email is sponsored by: Discover Easy Linux Migration Strategies
from IBM. Find simple to follow Roadmaps, straightforward articles,
informative Webcasts and more! Get everything you need to get up to
speed, fast. http://ads.osdn.com/?ad_id=7477&alloc_id=16492&op=click


<Prev in Thread] Current Thread [Next in Thread>