|
[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index] Re: iSCSI: Data Digest Error questionJulian Satran wrote: > > Yes there is a typo (the second appearance is yes). > However the function is correctly stated as OR i.e. the result is yes > if at least one of them wants order (it requires both to agree to say > no). Yes, you are right. I guess I was thinking negative logic. Thanks. -- Mallikarjun Mallikarjun Chadalapaka Networked Storage Architecture Network Storage Solutions Organization MS 5668 Hewlett-Packard, Roseville. cbm@rose.hp.com > > Julo > > "Mallikarjun C." > <cbm@rose.hp.com> To: ips > Sent by: owner-ips@ece.cmu.edu <ips@ece.cmu.edu> > cc: > 04-10-01 21:35 Subject: Re: Data > Please respond to cbm Digest Error question > > > > "Mallikarjun C." wrote: > > > > Prasenjit, > > > > The Reject mechanism is intended to be uniformly applicable to > > all (recovery-savvy and otherwise) targets. > > > > Recovery R2T, OTOH, is only for ErrorRecoveryLevel>=1 capable > > implementations (DataSequenceInOrder=true > > Sorry, that was supposed to be DataSequenceInOrder=no. > > Also, it appears to me that the result function for this key > should be AND, and there's a typo in the description - it says > "no" for both cases. Julian, comments? > > >as well, I think this > > needs to be called out in its description), since a recovery R2T > > implies going back in the data stream. > > > > The delivery subsystem failure status is the final termination > > status (even though there were several good data PDUs after the > > failed one) for ErrorRecoveryLevel=0 implementations, sent after > > waiting for the F-bit. > > > > Hope that helps. > > > > Regards. > > -- > > Mallikarjun > > > > Mallikarjun Chadalapaka > > Networked Storage Architecture > > Network Storage Solutions Organization > > MS 5668 Hewlett-Packard, Roseville. > > cbm@rose.hp.com > > > > Prasenjit Sarkar wrote: > > > > > > This question relates to how the target handles data digest errors > for data > > > pdus. > > > > > > The spec says the target has to send a reject pdu with the > attached pdu > > > header > > > + (recovery R2T OR delivery subsystem failure status) > > > > > > Why do we need two mechanisms? Wouldnt the pdu header indicate > what needed > > > to be transmitted? Maybe a word of motivation might help. > > > > > > Prasenjit Sarkar > > > Research Staff Member > > > IBM Almaden Research > > > San Jose > > -- > Mallikarjun > > Mallikarjun Chadalapaka > Networked Storage Architecture > Network Storage Solutions Organization > MS 5668 Hewlett-Packard, Roseville. > cbm@rose.hp.com
Home Last updated: Fri Oct 05 15:17:20 2001 7077 messages in chronological order |