|
[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index] RE: [Ips] Re: iscsi RDMA standardThe specs for iSCSI Extensions for RDMA (iSER) and the underlying = Datamover Architecture (DA) were submitted to the IETF in July. The = URLs for the Internet-Drafts are=20 http://www.ietf.org/internet-drafts/draft-chadalapaka-iwarp-da-00.txt http://www.ietf.org/internet-drafts/draft-ko-iwarp-iser-00.txt -----Original Message----- From: Julian Satran [mailto:Julian_Satran@il.ibm.com] Sent: Sunday, September 28, 2003 10:04 PM To: Vysyaraju Daliraju Cc: Eddy Quicksall; ips@ietf.org; ips-admin@ietf.org Subject: [Ips] Re: iscsi RDMA standard There is an extension to iSCSI called ISER - being worked on by a = consortium of companies. Julo=20 Vysyaraju Daliraju <VDaliraju@sbs.com>=20 28/09/2003 21:51 ToJulian Satran/Haifa/IBM@IBMIL, Eddy Quicksall = <eddy_quicksall@ivivity.com>=20 ccips@ietf.org, ips-admin@ietf.org=20 Subjectiscsi RDMA standard Hi=20 =20 I am new to RDMA ISCSI. Is there any standars exist for iscsi RDMA.=20 =20 Vysyaraju Dali Raju=20 __________________________________________=20 Vysyaraju Daliraju, / Sr Systems Software Engineer=20 SBS Technologies, 8371C Central Ave. / Newark, CA 94560=20 Phone: 510-742-2570 / Fax: -2501 / EMail: VDaliraju@sbs.com = <mailto:VDaliraju@sbs.com>=20 =20 =20 =20 =20 -----Original Message----- From: Julian Satran [mailto:Julian_Satran@il.ibm.com] Sent: Saturday, September 20, 2003 1:52 PM To: Eddy Quicksall Cc: ips@ietf.org; ips-admin@ietf.org Subject: RE: [Ips] NOP during Discovery Eddy,=20 I think that they realized that already and are content. NOP is somewhat = richer in that in a "storage router" NOP might reach the end-point while = SendTargets will probably not. In any case the text is unequivocal = about NOP.=20 Julo=20 Eddy Quicksall <eddy_quicksall@ivivity.com>=20 20/09/2003 14:12=20 ToJulian Satran/Haifa/IBM@IBMIL=20 ccips@ietf.org, ips-admin@ietf.org=20 SubjectRE: [Ips] NOP during Discovery Thanks=20 =20 Would it work to just use SendTargets=3D<nothing> instead of NOP? If = there is a target for the discovery session, it would respond with just = the target name. If there are no targets for the discovery session, the = target would still respond but with 0 for the DataSegmentLength.=20 =20 If so, can you talk that person into that?=20 =20 Eddy=20 =20 -----Original Message----- From: Julian Satran [mailto:Julian_Satran@il.ibm.com]=20 Sent: Saturday, September 20, 2003 1:52 PM To: Eddy Quicksall Cc: ips@ietf.org; ips-admin@ietf.org Subject: Re: [Ips] NOP during Discovery=20 =20 It is not allowed in discovery - There was a long discussion with one = party insisting on having NOP as a way of keeping the discovery session = alive.=20 Discovery being only a lightweight location mechanism it was decided = that it is not worth having it as a permanent alternative for a = management interface.=20 Regards,=20 Julo=20 Eddy Quicksall <eddy_quicksall@ivivity.com>=20 Sent by: ips-admin@ietf.org=20 19/09/2003 16:04=20 Toips@ietf.org=20 cc =20 Subject[Ips] NOP during Discovery =20 =20 I have found that at least one initiator is doing a NOP-out during = discovery phase. But the spec says that isn't allowed. Am I = misunderstanding something?=20 =20 3.3 iSCSI Session Types=20 iSCSI defines two types of sessions:=20 a) Normal operational session - an unrestricted session.=20 b) Discovery-session - a session only opened for target=20 discovery. The target MUST ONLY accept text requests with the=20 SendTargets key and a logout request with reason "close the=20 session". All other requests MUST be rejected.=20 The session type is defined during login with key=3Dvalue parameter = in=20 the login command.=20 =20 =20 Eddy=20 =20 ***This message has been scanned for virus, spam, and undesirable = content.*** ***For further information, contact your mail administrator.***=20 _______________________________________________ Ips mailing list Ips@ietf.org https://www1.ietf.org/mailman/listinfo/ips
Home Last updated: Mon Sep 29 14:19:44 2003 12913 messages in chronological order |