|
[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index] Re: Unsolicited Data.It says "all the unsolicited data associated ". You read it as "all the data associated". Does that help the confusion? Eddy ----- Original Message ----- From: "BURBRIDGE,MATTHEW (HP-UnitedKingdom,ex2)" <matthew_burbridge@hp.com> To: <ips@ece.cmu.edu> Sent: Thursday, June 07, 2001 8:33 AM Subject: RE: Unsolicited Data. > Hi Sanjeev, > > Perhaps I'm getting confused here: From 2.3.1 below, if the F bit is set > then it indicates the immediate data that acompany the command is all the > data associated with the command. However, if there is no immediated data, > but data is to be written the F bit can not be set to 1. In this case the > data is either sent a unsolicited Data PDUs or as solicited data PDUs. This > reverts back to my original point of changing the spec to state that if > unsolicited data has been negoiated then the initiator MUST honour it. > > With regards to your comment: "Is it still going to help if Unsolicited data > is not chosen to be sent while IntialR2t=No has been negotiated???? Or do I > have misunderstanding with a general delivery path here." may I can explain > my concern: If the initiator does not choose to send unsolicited data > although it has negoiated to do so, how is the data transfers insigated as > the target is expecting unsoliciated Data PDU but the initiator is expecting > an R2T? > > Feel free to call me so we can discuss offline. > > Thanks > > Matthew Burbridge > NIS-Bristol > Hewlett Packard > Telnet: 312 7010 > Tel UK (44) 117 312 7010 > E-mail: matthewb@bri.hp.com > > > > -----Original Message----- > From: Sanjeev Bhagat (TRIPACE/Zoetermeer) [mailto:sbhagat@tripace.com] > Sent: Thursday, June 07, 2001 1:09 PM > To: 'BURBRIDGE,MATTHEW (HP-UnitedKingdom,ex2)'; ips@ece.cmu.edu > Cc: 'steph@cs.uchicago.edu' > Subject: RE: Unsolicited Data. > > > Matthew, > > among the two options you gave I would like to comment on the second > > It is possible for initiator **not** to send any unsolicited data and send > the rest of data as solicited data by the use of bit 7 in Flags and Task > Attributes. This is posted here http://ips.pdl.cs.cmu.edu/mail/msg04647.html > > The new 2.3.1 is: > > 2.3.1 Flags and Task Attributes > > The flags for a SCSI Command are: > > b7 (F) set to 1 when the immediate data that accompany the command > are all the unsolicited data associated with this command. For a > write, if Expected Data Transfer Length is larger than the Length the > target may solicit additional data through R2T. > > About the option no. 1 Stephen has already explained. Refer to the very > specific paragraph below of his email > > In a hardware accelerate implementation, unsolicited data will be > handled differently than solicited data. The target can chose where > the solicited data is delivered, but unsolicited data arrives through > a general-delivery path. Therefore, if the offered unsolicited data > burst size approaches the regular burst size, an implementation will > need TWICE as much buffer memory (half in the general-delivery area > and half in the flow-controlled buffer area) to support the same > operation load. Maybe this tradeoff is OK, but given the cost > sensitivity of storage targets, I doubt it. > > I would like to comment one thing here that in the current iSCSI specs > implementation the complete data for a specific command Including (Cmd PDUS > , Data PDUs and Response PDUs) follow the same path. So the whole of data > will follow the general delivery path (same connection) whether solicited or > unsolicited. Is it still going to help if Unsolicited data is not chosen to > be sent while IntialR2t=No has been negotiated???? Or do I have > misunderstanding with a general delivery path here. > > Regards, > Sanjeev > > -----Original Message----- > From: BURBRIDGE,MATTHEW (HP-UnitedKingdom,ex2) > [mailto:matthew_burbridge@hp.com] > Sent: Thursday, June 07, 2001 12:40 AM > To: ips@ece.cmu.edu > Subject: RE: Unsolicited Data. > > > Hi Stephen, Sanjeev, and other interested parties, > > I whole heartily agree with both of your comments and see the benefit of > unsolicited data in removing the R2T induced latency for both small writes > and large writes on systems where such latency is an issue. However, for > large writes the unsolicited portion of the data need not be all the data in > the transfer but enough to overcome the initial latency. > > My main concern is that although the initiator and target have negoiated to > use unsolicited data (InitialR2t=no), in the spec it appears to be optional. > Therefore how does the target know that unsolicited data will be sent. > Currently it does not and the target will have to set a timer that expires > if no unsolicited data is forthcoming which will incur an even more > unacceptable latency. > > I see that there are two options: > 1)State in the spec that if unsolicited data has been negotiated then the > initiator MUST honour it. (My preferred) > 2)Have a flag in the SCSI Command PDU indicating that data is expected (No > thanks). > > Thanks > > Matthew Burbridge > NIS-Bristol > Hewlett Packard > Telnet: 312 7010 > E-mail: matthewb@bri.hp.com > > > -----Original Message----- > From: Stephen Bailey [mailto:steph@cs.uchicago.edu] > Sent: Thursday, June 07, 2001 2:20 AM > To: ips@ece.cmu.edu > Subject: Re: Unsolicited Data. > > > Sanjeev, > > > why would initiator choose not to send unsolicited immediate data when > > it has negotiated for IntialR2T= Yes?? > > I assume you mean InitialR2T=no. > > I would do this because I think that the real benefit of unsolicited > data is for writes where ALL the write data can be sent unsolicited, > and I'm happy to have the target schedule the transfer of (ALL) data > for larger writes. Of course, others would say I'm a moron. > > Assuming the overwhelming majority of storage use is file access, > small writes are exclusively file system metadata writes. Unsolicited > data is a huge win for metadata writes because the latency of these > metadata writes is frequently not covered by transferring other data > (within a single access thread). On the other hand, the file system > (usually) generates enough concurrent demand for real data writes that > the flow control (R2T) latency is well hidden. > > You might argue that if you're running a file system optimized fors > today's SANs over a link with a HUGE bandwidth * delay product > (greater than the expected demand created by the file system), sending > large amounts of unsolicited write data will substantially improve the > link utilization. That is true, but I don't think many targets are > going to offer unsolicited data bursts remotely that large. > > In a hardware accelerate implementation, unsolicited data will be > handled differently than solicited data. The target can chose where > the solicited data is delivered, but unsolicited data arrives through > a general-delivery path. Therefore, if the offered unsolicited data > burst size approaches the regular burst size, an implementation will > need TWICE as much buffer memory (half in the general-delivery area > and half in the flow-controlled buffer area) to support the same > operation load. Maybe this tradeoff is OK, but given the cost > sensitivity of storage targets, I doubt it. > > As always, I welcome comments from those who see things differently. > > Steph >
Home Last updated: Tue Sep 04 01:04:32 2001 6315 messages in chronological order |