draft-ietf-nfsv4-nfs-ulb-v2-06.txt   draft-ietf-nfsv4-nfs-ulb-v2-07.txt 
Network File System Version 4 C. Lever Network File System Version 4 C. Lever
Internet-Draft Oracle Internet-Draft Oracle
Intended status: Standards Track 16 November 2021 Intended status: Standards Track 13 May 2022
Expires: 20 May 2022 Expires: 14 November 2022
Network File System (NFS) Upper-Layer Binding To RPC-Over-RDMA Version 2 Network File System (NFS) Upper-Layer Binding To RPC-Over-RDMA Version 2
draft-ietf-nfsv4-nfs-ulb-v2-06 draft-ietf-nfsv4-nfs-ulb-v2-07
Abstract Abstract
This document specifies Upper-Layer Bindings of Network File System This document specifies Upper-Layer Bindings of Network File System
(NFS) protocol versions to RPC-over-RDMA version 2. (NFS) protocol versions to RPC-over-RDMA version 2.
Note Note
Discussion of this draft takes place on the NFSv4 working group Discussion of this draft takes place on the NFSv4 working group
mailing list (nfsv4@ietf.org), archived at mailing list (nfsv4@ietf.org), archived at
skipping to change at page 1, line 43 skipping to change at page 1, line 43
Internet-Drafts are working documents of the Internet Engineering Internet-Drafts are working documents of the Internet Engineering
Task Force (IETF). Note that other groups may also distribute Task Force (IETF). Note that other groups may also distribute
working documents as Internet-Drafts. The list of current Internet- working documents as Internet-Drafts. The list of current Internet-
Drafts is at https://datatracker.ietf.org/drafts/current/. Drafts is at https://datatracker.ietf.org/drafts/current/.
Internet-Drafts are draft documents valid for a maximum of six months Internet-Drafts are draft documents valid for a maximum of six months
and may be updated, replaced, or obsoleted by other documents at any and may be updated, replaced, or obsoleted by other documents at any
time. It is inappropriate to use Internet-Drafts as reference time. It is inappropriate to use Internet-Drafts as reference
material or to cite them other than as "work in progress." material or to cite them other than as "work in progress."
This Internet-Draft will expire on 20 May 2022. This Internet-Draft will expire on 14 November 2022.
Copyright Notice Copyright Notice
Copyright (c) 2021 IETF Trust and the persons identified as the Copyright (c) 2022 IETF Trust and the persons identified as the
document authors. All rights reserved. document authors. All rights reserved.
This document is subject to BCP 78 and the IETF Trust's Legal This document is subject to BCP 78 and the IETF Trust's Legal
Provisions Relating to IETF Documents (https://trustee.ietf.org/ Provisions Relating to IETF Documents (https://trustee.ietf.org/
license-info) in effect on the date of publication of this document. license-info) in effect on the date of publication of this document.
Please review these documents carefully, as they describe your rights Please review these documents carefully, as they describe your rights
and restrictions with respect to this document. Code Components and restrictions with respect to this document. Code Components
extracted from this document must include Simplified BSD License text extracted from this document must include Revised BSD License text as
as described in Section 4.e of the Trust Legal Provisions and are described in Section 4.e of the Trust Legal Provisions and are
provided without warranty as described in the Simplified BSD License. provided without warranty as described in the Revised BSD License.
Table of Contents Table of Contents
1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . 3 1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . 3
2. Requirements Language . . . . . . . . . . . . . . . . . . . . 3 2. Requirements Language . . . . . . . . . . . . . . . . . . . . 3
3. Upper-Layer Binding for NFS Versions 2 and 3 . . . . . . . . 4 3. Upper-Layer Binding for NFS Versions 2 and 3 . . . . . . . . 4
3.1. Reply Size Estimation . . . . . . . . . . . . . . . . . . 4 3.1. DDP-Eligibility . . . . . . . . . . . . . . . . . . . . . 4
3.2. RPC Binding Considerations . . . . . . . . . . . . . . . 5 3.2. Reply Size Estimation . . . . . . . . . . . . . . . . . . 5
3.3. Transport Considerations . . . . . . . . . . . . . . . . 5 3.3. RPC Binding Considerations . . . . . . . . . . . . . . . 5
3.3.1. Keep-Alive . . . . . . . . . . . . . . . . . . . . . 5 3.4. Transport Considerations . . . . . . . . . . . . . . . . 5
3.3.2. Replay Detection . . . . . . . . . . . . . . . . . . 6 3.4.1. Keep-Alive . . . . . . . . . . . . . . . . . . . . . 6
3.4.2. Replay Detection . . . . . . . . . . . . . . . . . . 6
4. Upper-Layer Bindings for NFS Version 2 and 3 Auxiliary 4. Upper-Layer Bindings for NFS Version 2 and 3 Auxiliary
Protocols . . . . . . . . . . . . . . . . . . . . . . . . 6 Protocols . . . . . . . . . . . . . . . . . . . . . . . . 7
4.1. MOUNT, NLM, and NSM Protocols . . . . . . . . . . . . . . 7 4.1. MOUNT, NLM, and NSM Protocols . . . . . . . . . . . . . . 7
4.2. NFSACL Protocol . . . . . . . . . . . . . . . . . . . . . 7 4.2. NFSACL Protocol . . . . . . . . . . . . . . . . . . . . . 7
5. Upper-Layer Binding For NFS Version 4 . . . . . . . . . . . . 7 5. Upper-Layer Binding For NFS Version 4 . . . . . . . . . . . . 8
5.1. DDP-Eligibility . . . . . . . . . . . . . . . . . . . . . 7 5.1. DDP-Eligibility . . . . . . . . . . . . . . . . . . . . . 8
5.1.1. The NFSv4.2 READ_PLUS operation . . . . . . . . . . . 8 5.1.1. The NFSv4.2 READ_PLUS operation . . . . . . . . . . . 8
5.2. Reply Size Estimation . . . . . . . . . . . . . . . . . . 8 5.1.2. NFS Version 4 COMPOUND Requests . . . . . . . . . . . 9
5.2.1. Reply Size Estimation for Minor Version 0 . . . . . . 9 5.2. Reply Size Estimation . . . . . . . . . . . . . . . . . . 11
5.2.1. Reply Size Estimation for Minor Version 0 . . . . . . 11
5.2.2. Reply Size Estimation for Minor Version 1 and 5.2.2. Reply Size Estimation for Minor Version 1 and
Newer . . . . . . . . . . . . . . . . . . . . . . . . 9 Newer . . . . . . . . . . . . . . . . . . . . . . . . 12
5.3. RPC Binding Considerations . . . . . . . . . . . . . . . 9 5.3. RPC Binding Considerations . . . . . . . . . . . . . . . 12
5.4. NFS COMPOUND Requests . . . . . . . . . . . . . . . . . . 9 5.4. Transport Considerations . . . . . . . . . . . . . . . . 12
5.4.1. Multiple DDP-eligible Data Items . . . . . . . . . . 10 5.4.1. Congestion Avoidance . . . . . . . . . . . . . . . . 12
5.4.2. Chunk List Complexity . . . . . . . . . . . . . . . . 10 5.4.2. Retransmission and Keep-alive . . . . . . . . . . . . 13
5.4.3. NFS Version 4 COMPOUND Example . . . . . . . . . . . 11 5.5. Session-Related Considerations . . . . . . . . . . . . . 13
5.5. NFS Callback Requests . . . . . . . . . . . . . . . . . . 11 6. Upper-Layer Binding For NFS Version 4 Callbacks . . . . . . . 14
5.5.1. NFS Version 4.0 Callback . . . . . . . . . . . . . . 12 6.1. NFS Version 4.0 Callback . . . . . . . . . . . . . . . . 14
5.5.2. NFS Version 4.1 Callback . . . . . . . . . . . . . . 12 6.2. NFS Version 4.1 Callback . . . . . . . . . . . . . . . . 15
5.6. Session-Related Considerations . . . . . . . . . . . . . 12 7. Extending NFS Upper-Layer Bindings . . . . . . . . . . . . . 15
5.7. Transport Considerations . . . . . . . . . . . . . . . . 13 8. Security Considerations . . . . . . . . . . . . . . . . . . . 15
5.7.1. Congestion Avoidance . . . . . . . . . . . . . . . . 13 9. IANA Considerations . . . . . . . . . . . . . . . . . . . . . 16
5.7.2. Retransmission and Keep-alive . . . . . . . . . . . . 14 10. References . . . . . . . . . . . . . . . . . . . . . . . . . 16
6. Extending NFS Upper-Layer Bindings . . . . . . . . . . . . . 15 10.1. Normative References . . . . . . . . . . . . . . . . . . 16
7. Security Considerations . . . . . . . . . . . . . . . . . . . 15 10.2. Informative References . . . . . . . . . . . . . . . . . 17
8. IANA Considerations . . . . . . . . . . . . . . . . . . . . . 15 Acknowledgments . . . . . . . . . . . . . . . . . . . . . . . . . 18
9. References . . . . . . . . . . . . . . . . . . . . . . . . . 15 Author's Address . . . . . . . . . . . . . . . . . . . . . . . . 18
9.1. Normative References . . . . . . . . . . . . . . . . . . 15
9.2. Informative References . . . . . . . . . . . . . . . . . 16
Acknowledgments . . . . . . . . . . . . . . . . . . . . . . . . . 17
Author's Address . . . . . . . . . . . . . . . . . . . . . . . . 17
1. Introduction 1. Introduction
The RPC-over-RDMA version 2 transport can employ direct data The RPC-over-RDMA version 2 transport can employ direct data
placement to convey data payloads associated with RPC transactions, placement to convey data payloads associated with RPC transactions,
as described in [I-D.ietf-nfsv4-rpcrdma-version-two]. As mandated by as described in [I-D.ietf-nfsv4-rpcrdma-version-two]. As mandated by
that document, RPC client and server implementations using RPC-over- that document, RPC client and server implementations using RPC-over-
RDMA version 2 MUST agree in advance which XDR data items and RPC RDMA version 2 MUST agree in advance which XDR data items and RPC
procedures are eligible for direct data placement (DDP). procedures are eligible for direct data placement (DDP).
skipping to change at page 4, line 14 skipping to change at page 4, line 14
3. Upper-Layer Binding for NFS Versions 2 and 3 3. Upper-Layer Binding for NFS Versions 2 and 3
The Upper-Layer Binding specification in this section applies to NFS The Upper-Layer Binding specification in this section applies to NFS
version 2 [RFC1094] and NFS version 3 [RFC1813]. For brevity, in version 2 [RFC1094] and NFS version 3 [RFC1813]. For brevity, in
this document, a "Legacy NFS client" refers to an NFS client using this document, a "Legacy NFS client" refers to an NFS client using
version 2 or version 3 of the NFS RPC program (100003) to communicate version 2 or version 3 of the NFS RPC program (100003) to communicate
with an NFS server. Likewise, a "Legacy NFS server" is an NFS server with an NFS server. Likewise, a "Legacy NFS server" is an NFS server
communicating with clients using NFS version 2 or NFS version 3. communicating with clients using NFS version 2 or NFS version 3.
The following XDR data items in NFS versions 2 and 3 are DDP- 3.1. DDP-Eligibility
eligible:
Generally, storage protocols based on RDMA divide both read and write
operations into two steps. This division enables the payload
receiver to allocate the sink buffer for each I/O operation in
advance of the network payload transfer. By allocating the sink
buffer tactically, a good quality receiver implementation reduces the
amount of data movement it must perform during and after the I/O
operation.
During an NFS WRITE that involves explicit RDMA, first the NFS client
sends a request that indicates where the NFS server can find the
payload buffer, then the NFS server pulls the WRITE payload from that
buffer. Likewise, during an NFS READ that involves explicit RDMA,
the NFS client provides the location of the destination buffer, then
the NFS server pushes the READ payload to that buffer.
Therefore, the following XDR data items in NFS versions 2 and 3 are
DDP-eligible:
* The opaque file data argument in the NFS WRITE procedure * The opaque file data argument in the NFS WRITE procedure
* The pathname argument in the NFS SYMLINK procedure * The pathname argument in the NFS SYMLINK procedure
* The opaque file data result in the NFS READ procedure * The opaque file data result in the NFS READ procedure
* The pathname result in the NFS READLINK procedure * The pathname result in the NFS READLINK procedure
All other argument or result data items in NFS versions 2 and 3 are All other argument or result data items in NFS versions 2 and 3 are
not DDP-eligible. not DDP-eligible.
Regardless of whether an NFS operation is considered non-idempotent, Regardless of whether an NFS operation is considered non-idempotent,
a transport error might not indicate whether the server has processed a transport error might not indicate whether the server has processed
the arguments of the RPC Call or whether the server has accessed or the arguments of the RPC Call or whether the server has accessed or
modified client memory associated with that RPC. modified client memory associated with that RPC.
3.1. Reply Size Estimation 3.2. Reply Size Estimation
During the construction of each RPC Call message, a Requester is During the construction of each RPC Call message, a Requester is
responsible for allocating appropriate RDMA resources to receive the responsible for allocating appropriate RDMA resources to receive the
corresponding Reply message. These resources must be capable of corresponding Reply message. These resources must be capable of
holding the entire Reply. Therefore the Requester needs to estimate holding the entire Reply. Therefore the Requester needs to estimate
the maximum possible size of the expected Reply message. the maximum possible size of the expected Reply message.
* Often, the expected Reply can fit in a limited number of RDMA Send * Often, the expected Reply can fit in a limited number of RDMA Send
messages. The Requester need not provision any RDMA resources for messages. The Requester need not provision any RDMA resources for
the Reply, relying instead on message continuation to handle the the Reply, relying instead on message continuation to handle the
skipping to change at page 5, line 16 skipping to change at page 5, line 35
Reply chunk. The Requester MUST reliably estimate the maximum Reply chunk. The Requester MUST reliably estimate the maximum
size of the payload received via the Reply chunk. size of the payload received via the Reply chunk.
* If RDMA resources are not available to send a Reply, a Responder * If RDMA resources are not available to send a Reply, a Responder
falls back to message continuation. falls back to message continuation.
A correctly implemented Legacy NFS client thus avoids retransmission A correctly implemented Legacy NFS client thus avoids retransmission
of non-idempotent NFS requests due to improperly estimated Reply of non-idempotent NFS requests due to improperly estimated Reply
resources. resources.
3.2. RPC Binding Considerations 3.3. RPC Binding Considerations
Legacy NFS servers typically listen for clients on UDP and TCP port Legacy NFS servers typically listen for clients on UDP and TCP port
2049. Additionally, they register these ports with a local 2049. Additionally, they register these ports with a local
portmapper service [RFC1833]. portmapper service [RFC1833].
A Legacy NFS server supporting RPC-over-RDMA version 2 and A Legacy NFS server supporting RPC-over-RDMA version 2 and
registering itself with the RPC portmapper MAY choose an arbitrary registering itself with the RPC portmapper MAY choose an arbitrary
port or MAY use the alternative well-known port number for its RPC- port or MAY use the alternative well-known port number for its RPC-
over-RDMA service (see Section 8). The chosen port MAY be registered over-RDMA service (see Section 9). The chosen port MAY be registered
with the RPC portmapper using the netids assigned in Section 12 of with the RPC portmapper using the netids assigned in Section 12 of
[I-D.ietf-nfsv4-rpcrdma-version-two]. [I-D.ietf-nfsv4-rpcrdma-version-two].
3.3. Transport Considerations 3.4. Transport Considerations
3.4.1. Keep-Alive
3.3.1. Keep-Alive
Legacy NFS client implementations can rely on connection keep-alive Legacy NFS client implementations can rely on connection keep-alive
to detect when a Legacy NFS server has become unresponsive. When an to detect when a Legacy NFS server has become unresponsive. When an
NFS server is no longer responsive, client-side keep-alive terminates NFS server is no longer responsive, client-side keep-alive terminates
the connection, triggering reconnection and retransmission of the connection, triggering reconnection and retransmission of
outstanding RPC transactions. outstanding RPC transactions.
Some RDMA transports (such as the Reliable Connected QP type on Some RDMA transports (such as the Reliable Connected QP type on
InfiniBand) have no keep-alive mechanism. Without a disconnect or InfiniBand) have no keep-alive mechanism. Without a disconnect or
new RPC traffic, such connections can remain alive long after an NFS new RPC traffic, such connections can remain alive long after an NFS
skipping to change at page 6, line 5 skipping to change at page 6, line 26
has consumed all available RPC-over-RDMA version 2 credits on that has consumed all available RPC-over-RDMA version 2 credits on that
transport connection, it awaits a reply indefinitely before sending transport connection, it awaits a reply indefinitely before sending
another RPC request. another RPC request.
Legacy NFS clients SHOULD reserve one RPC-over-RDMA version 2 credit Legacy NFS clients SHOULD reserve one RPC-over-RDMA version 2 credit
to use for periodic server or connection health assessment. Either to use for periodic server or connection health assessment. Either
peer can use this credit to drive an RPC request on an otherwise idle peer can use this credit to drive an RPC request on an otherwise idle
connection, triggering either an affirmative server response or a connection, triggering either an affirmative server response or a
connection termination. connection termination.
3.3.2. Replay Detection 3.4.2. Replay Detection
Like NFSv4.0, Legacy NFS servers typically employ request replay Like NFSv4.0, Legacy NFS servers typically employ request replay
detection to reduce the risk of data and file namespace corruption detection to reduce the risk of data and file namespace corruption
that could result when an NFS client retransmits a non-idempotent NFS that could result when an NFS client retransmits a non-idempotent NFS
request. A Legacy NFS server can send a cached response when a request. A Legacy NFS server can send a cached response when a
replay is detected, rather than executing the request again. Replay replay is detected, rather than executing the request again. Replay
detection is not perfect, but it is usually adequate. detection is not perfect, but it is usually adequate.
For Legacy NFS servers, replay detection commonly utilizes heuristic For Legacy NFS servers, replay detection commonly utilizes heuristic
indicators such as the IP address of the NFS client, the source port indicators such as the IP address of the NFS client, the source port
skipping to change at page 8, line 41 skipping to change at page 9, line 17
to predict the size or structure of a READ_PLUS Reply in advance. to predict the size or structure of a READ_PLUS Reply in advance.
The use of direct data placement is therefore challenging. Moreover, The use of direct data placement is therefore challenging. Moreover,
the usual benefits of hardware-assisted data placement are entirely the usual benefits of hardware-assisted data placement are entirely
lost if the client must parse the result of each READ I/O. lost if the client must parse the result of each READ I/O.
Therefore this Upper Layer Binding does not make elements of an Therefore this Upper Layer Binding does not make elements of an
NFSv4.2 READ_PLUS Reply DDP-eligible. Further, this Upper Layer NFSv4.2 READ_PLUS Reply DDP-eligible. Further, this Upper Layer
Binding recommends that NFS client implemenations avoid using the Binding recommends that NFS client implemenations avoid using the
READ_PLUS operation on NFS/RDMA mount points. READ_PLUS operation on NFS/RDMA mount points.
5.2. Reply Size Estimation 5.1.2. NFS Version 4 COMPOUND Requests
Within NFS version 4, there are certain variable-length result data
items whose maximum size cannot be estimated by clients reliably
because there is no protocol-specified size limit on these result
arrays. These include:
* The attrlist4 field
* Fields containing ACLs such as fattr4_acl, fattr4_dacl, and
fattr4_sacl
* Fields in the fs_locations4 and fs_locations_info4 data structures
* Fields which pertain to pNFS layout metadata, such as loc_body,
loh_body, da_addr_body, lou_body, lrf_body, fattr_layout_types,
and fs_layout_types
5.2.1. Reply Size Estimation for Minor Version 0
The NFS version 4.0 protocol itself does not impose any bound on the
size of NFS Calls or Replies.
Variable-length fattr4 attributes make it particularly difficult for
clients to predict the maximum size of some NFS version 4.0 Replies.
Client implementations might rely upon internal architectural limits
to constrain the reply size, but such limits are not always reliable.
When an NFS version 4.0 client cannot predict the size of a Reply, it
can rely on message continuation to enable a Reply under any
circumstances.
5.2.2. Reply Size Estimation for Minor Version 1 and Newer
In NFS version 4.1 and newer minor versions, the csa_fore_chan_attrs
argument of the CREATE_SESSION operation contains a
ca_maxresponsesize field. The value in this field is the absolute
maximum size of replies generated by an NFS version 4.1 server.
An NFS version 4 client can use this value when it is impossible to
estimate a reply size upper bound precisely. In practice, objects
such as ACLs, named attributes, layout bodies, and security labels
are much smaller than this maximum.
5.3. RPC Binding Considerations
NFS version 4 servers are required to listen on TCP port 2049 and are
not required to register with an rpcbind service [RFC7530].
Therefore, an NFS version 4 server supporting RPC-over-RDMA version 2
MUST use the alternative well-known port number for its RPC-over-RDMA
service defined in Section 8.
5.4. NFS COMPOUND Requests 5.1.2.1. Multiple DDP-eligible Data Items
5.4.1. Multiple DDP-eligible Data Items
An NFS version 4 COMPOUND procedure can contain more than one An NFS version 4 COMPOUND procedure can contain more than one
operation that carries a DDP-eligible data item. An NFS version 4 operation that carries a DDP-eligible data item. An NFS version 4
client provides XDR Position values in each Read chunk to determine client provides XDR Position values in each Read chunk to determine
which chunk is associated with which argument data item. However, which chunk is associated with which argument data item. However,
NFS version 4 server and client implementations must agree on how to NFS version 4 server and client implementations must agree on how to
pair Write chunks with returned result data items. pair Write chunks with returned result data items.
A "READ operation" refers to any NFS version 4 operation with a DDP- A "READ operation" refers to any NFS version 4 operation with a DDP-
eligible result data item in the following lists. An NFS version 4 eligible result data item in the following lists. An NFS version 4
skipping to change at page 10, line 45 skipping to change at page 10, line 14
* If a READ operation returns a union arm which does not contain a * If a READ operation returns a union arm which does not contain a
DDP-eligible result, and the NFS version 4 client has provided a DDP-eligible result, and the NFS version 4 client has provided a
matching non-empty Write chunk, an NFS version 4 server MUST matching non-empty Write chunk, an NFS version 4 server MUST
return an empty Write chunk in that Write list position. return an empty Write chunk in that Write list position.
* If there are more READ operations than Write chunks, then * If there are more READ operations than Write chunks, then
remaining NFS Read operations in an NFS version 4 COMPOUND that remaining NFS Read operations in an NFS version 4 COMPOUND that
have no matching Write chunk MUST return their results inline. have no matching Write chunk MUST return their results inline.
5.4.2. Chunk List Complexity 5.1.2.2. Chunk List Complexity
By default, the RPC-over-RDMA version 2 protocol limits the number of By default, the RPC-over-RDMA version 2 protocol limits the number of
chunks or segments that may appear in Read or Write lists (see chunks or segments that may appear in Read or Write lists (see
Section 5.2 of [I-D.ietf-nfsv4-rpcrdma-version-two]). Section 5.2 of [I-D.ietf-nfsv4-rpcrdma-version-two]).
These implementation limits are significant when Kerberos integrity These implementation limits are significant when Kerberos integrity
or privacy is in use [RFC7861]. GSS services increase the size of or privacy is in use [RFC7861]. GSS services increase the size of
credential material in RPC headers, potentially requiring the more credential material in RPC headers, potentially requiring the more
frequent use of less efficient Special Payload or Continued Payload frequent use of less efficient Special Payload or Continued Payload
messages. messages.
skipping to change at page 11, line 26 skipping to change at page 10, line 41
* The Read list can contain either a Call chunk, no more than one * The Read list can contain either a Call chunk, no more than one
Read chunk, or both a Call chunk and one Read chunk. Read chunk, or both a Call chunk and one Read chunk.
* The Write list can contain no more than one Write chunk. * The Write list can contain no more than one Write chunk.
NFS version 4 clients wishing to send more complex chunk lists can NFS version 4 clients wishing to send more complex chunk lists can
use transport properties to bound the complexity of NFS version 4 use transport properties to bound the complexity of NFS version 4
COMPOUNDs, limit the number of elements in scatter-gather operations, COMPOUNDs, limit the number of elements in scatter-gather operations,
and avoid other sources of chunk overruns at the receiving peer. and avoid other sources of chunk overruns at the receiving peer.
5.4.3. NFS Version 4 COMPOUND Example 5.1.2.3. NFS Version 4 COMPOUND Example
The following example shows a Write list with three Write chunks, A, The following example shows a Write list with three Write chunks, A,
B, and C. The NFS version 4 server consumes the provided Write B, and C. The NFS version 4 server consumes the provided Write
chunks by writing the results of the designated operations in the chunks by writing the results of the designated operations in the
compound request (READ and READLINK) back to each chunk. compound request (READ and READLINK) back to each chunk.
Write list: Write list:
A --> B --> C A --> B --> C
skipping to change at page 11, line 48 skipping to change at page 11, line 20
PUTFH LOOKUP READ PUTFH LOOKUP READLINK PUTFH LOOKUP READ PUTFH LOOKUP READ PUTFH LOOKUP READLINK PUTFH LOOKUP READ
| | | | | |
v v v v v v
A B C A B C
If the NFS version 4 client does not want the READLINK result If the NFS version 4 client does not want the READLINK result
returned via RDMA, it provides an empty Write chunk for buffer B to returned via RDMA, it provides an empty Write chunk for buffer B to
indicate that the READLINK result must be returned inline. indicate that the READLINK result must be returned inline.
5.5. NFS Callback Requests 5.2. Reply Size Estimation
The NFS version 4 family of protocols supports server-initiated
callbacks to notify NFS version 4 clients of events such as recalled
delegations.
5.5.1. NFS Version 4.0 Callback Within NFS version 4, there are certain variable-length result data
items whose maximum size cannot be estimated by clients reliably
because there is no protocol-specified size limit on these result
arrays. These include:
An NFS version 4.0 client uses the SETCLIENTID operation for * The attrlist4 field
advertising the IP address, port, and netid of its NFS version 4.0
callback service. When an NFS version 4.0 server provides a
backchannel service to an NFS version 4.0 client that uses RPC-over-
RDMA version 2 for its forward channel, the server MUST advertise the
backchannel service using either the "tcp" or "tcp6" netid.
Because the NFSv4.0 backchannel does not operate on RPC-over-RDMA, * Fields containing ACLs such as fattr4_acl, fattr4_dacl, and
this document does not specify an Upper-Layer binding for the NFSv4.0 fattr4_sacl
backchannel RPC program.
5.5.2. NFS Version 4.1 Callback * Fields in the fs_locations4 and fs_locations_info4 data structures
In NFS version 4.1 and newer minor versions, callback operations may * Fields which pertain to pNFS layout metadata, such as loc_body,
appear on the same connection that is in use for NFS version 4 loh_body, da_addr_body, lou_body, lrf_body, fattr_layout_types,
forward channel client requests. NFS version 4 clients and servers and fs_layout_types
MUST use the mechanisms described in Section 4.5 of
[I-D.ietf-nfsv4-rpcrdma-version-two] to convey backchannel operations
on an RPC-over-RDMA version 2 transport.
The csa_back_chan_attrs argument of the CREATE_SESSION operation 5.2.1. Reply Size Estimation for Minor Version 0
contains a ca_maxresponsesize field. The value in this field is the
absolute maximum size of backchannel replies generated by a replying
NFS version 4 client.
There are no DDP-eligible data items in callback procedures defined The NFS version 4.0 protocol itself does not impose any bound on the
in NFS version 4.1 or NFS version 4.2. However, some callback size of NFS Calls or Replies.
operations, such as messages that convey device ID information, can
be sizeable. A sender can use Message Continuation or a Special
Payload message in this situation.
When an NFS version 4.1 client can support Special Payload Calls in Variable-length fattr4 attributes make it particularly difficult for
its backchannel, it reports a backchannel ca_maxrequestsize that is clients to predict the maximum size of some NFS version 4.0 Replies.
larger than the connection's inline thresholds. Otherwise, an NFS Client implementations might rely upon internal architectural limits
version 4 server MUST use only Simple Payload or Continued Payload to constrain the reply size, but such limits are not always reliable.
messages to convey backchannel operations. When an NFS version 4.0 client cannot predict the size of a Reply, it
can rely on message continuation to enable a Reply under any
circumstances.
5.6. Session-Related Considerations 5.2.2. Reply Size Estimation for Minor Version 1 and Newer
The presence of an NFS version 4 session (as defined in [RFC8881]) In NFS version 4.1 and newer minor versions, the csa_fore_chan_attrs
does not affect the operation of RPC-over-RDMA version 2. None of argument of the CREATE_SESSION operation contains a
the operations introduced to support NFS sessions (e.g., the SEQUENCE ca_maxresponsesize field. The value in this field is the absolute
operation) contain DDP-eligible data items. There is no need to maximum size of replies generated by an NFS version 4.1 server.
match the number of session slots with the available RPC-over-RDMA
version 2 credits.
However, there are a few new cases where an RPC transaction can fail. An NFS version 4 client can use this value when it is impossible to
For example, a Requester might receive, in response to an RPC estimate a reply size upper bound precisely. In practice, objects
request, an RDMA2_ERROR message with a rdma_err value of such as ACLs, named attributes, layout bodies, and security labels
RDMA2_ERR_BADXDR. These situations are not different from existing are much smaller than this maximum.
RPC errors, which an NFS session implementation can already handle
for other transport types. Moreover, there might be no SEQUENCE
result available to the Requester to distinguish whether failure
occurred before or after the Responder executed the requested
operations.
When a transport error occurs (e.g., an RDMA2_ERROR type message is 5.3. RPC Binding Considerations
received), the Requester proceeds, as usual, to match the incoming
XID value to a waiting RPC Call. The Requester terminates the RPC
transaction and reports the result status to the RPC consumer. The
Requester's session implementation then determines the session ID and
slot for the failed request and performs slot recovery to make that
slot usable again. Otherwise, that slot is rendered permanently
unavailable.
When an NFS session is not present (for example, when NFS version 4.0 NFS version 4 servers are required to listen on TCP port 2049 and are
is in use), a transport error does not indicate whether the server not required to register with an rpcbind service [RFC7530].
has processed the arguments of the RPC Call, or whether the server Therefore, an NFS version 4 server supporting RPC-over-RDMA version 2
has accessed or modified client memory associated with that RPC. MUST use the alternative well-known port number for its RPC-over-RDMA
service defined in Section 9.
5.7. Transport Considerations 5.4. Transport Considerations
5.7.1. Congestion Avoidance 5.4.1. Congestion Avoidance
Section 3.1 of [RFC7530] states: Section 3.1 of [RFC7530] states:
Where an NFS version 4 implementation supports operation over the Where an NFS version 4 implementation supports operation over the
IP network protocol, the supported transport layer between NFS and IP network protocol, the supported transport layer between NFS and
IP MUST be an IETF standardized transport protocol that is IP MUST be an IETF standardized transport protocol that is
specified to avoid network congestion; such transports include TCP specified to avoid network congestion; such transports include TCP
and the Stream Control Transmission Protocol (SCTP). and the Stream Control Transmission Protocol (SCTP).
Section 2.9.1 of [RFC8881] further states: Section 2.9.1 of [RFC8881] further states:
skipping to change at page 14, line 10 skipping to change at page 13, line 5
NFS version 4.1; however, reliable and in-order delivery of data NFS version 4.1; however, reliable and in-order delivery of data
combined with congestion control by the connectionless transport combined with congestion control by the connectionless transport
is REQUIRED. As a consequence, UDP by itself MUST NOT be used as is REQUIRED. As a consequence, UDP by itself MUST NOT be used as
an NFS version 4.1 transport. an NFS version 4.1 transport.
RPC-over-RDMA version 2 utilizes only reliable, connection-oriented RPC-over-RDMA version 2 utilizes only reliable, connection-oriented
transports that guarantee in-order delivery, meeting all the above transports that guarantee in-order delivery, meeting all the above
requirements for NFS version 4.0 and 4.1. See Section 4.2.1 of requirements for NFS version 4.0 and 4.1. See Section 4.2.1 of
[I-D.ietf-nfsv4-rpcrdma-version-two] for more details. [I-D.ietf-nfsv4-rpcrdma-version-two] for more details.
5.7.2. Retransmission and Keep-alive 5.4.2. Retransmission and Keep-alive
NFS version 4 client implementations often rely on a transport-layer NFS version 4 client implementations often rely on a transport-layer
connection keep-alive mechanism to detect when an NFS version 4 connection keep-alive mechanism to detect when an NFS version 4
server has become unresponsive. When an NFS server is no longer server has become unresponsive. When an NFS server is no longer
responsive, client-side keep-alive terminates the connection, responsive, client-side keep-alive terminates the connection,
triggering reconnection and RPC retransmission. triggering reconnection and RPC retransmission.
Some RDMA transports (such as the Reliable Connected QP type on Some RDMA transports (such as the Reliable Connected QP type on
InfiniBand) have no keep-alive mechanism. Without a disconnect or InfiniBand) have no keep-alive mechanism. Without a disconnect or
new RPC traffic, such connections can remain alive long after an NFS new RPC traffic, such connections can remain alive long after an NFS
skipping to change at page 15, line 5 skipping to change at page 13, line 42
* If a transport error occurs (e.g., an RDMA2_ERROR type message is * If a transport error occurs (e.g., an RDMA2_ERROR type message is
received) just before the disconnect or instead of a disconnect, received) just before the disconnect or instead of a disconnect,
the Requester MUST respond to that error as prescribed by the the Requester MUST respond to that error as prescribed by the
specification of the RPC transport. Then the NFS version 4 rules specification of the RPC transport. Then the NFS version 4 rules
for handling retransmission apply. for handling retransmission apply.
* If there is a transport disconnect and the Responder has provided * If there is a transport disconnect and the Responder has provided
no other response for a request, then only the NFS version 4 rules no other response for a request, then only the NFS version 4 rules
for handling retransmission apply. for handling retransmission apply.
6. Extending NFS Upper-Layer Bindings 5.5. Session-Related Considerations
The presence of an NFS version 4 session (as defined in [RFC8881])
does not affect the operation of RPC-over-RDMA version 2. None of
the operations introduced to support NFS sessions (e.g., the SEQUENCE
operation) contain DDP-eligible data items. There is no need to
match the number of session slots with the available RPC-over-RDMA
version 2 credits.
However, there are a few new cases where an RPC transaction can fail.
For example, a Requester might receive, in response to an RPC
request, an RDMA2_ERROR message with a rdma_err value of
RDMA2_ERR_BADXDR. These situations are not different from existing
RPC errors, which an NFS session implementation can already handle
for other transport types. Moreover, there might be no SEQUENCE
result available to the Requester to distinguish whether failure
occurred before or after the Responder executed the requested
operations.
When a transport error occurs (e.g., an RDMA2_ERROR type message is
received), the Requester proceeds, as usual, to match the incoming
XID value to a waiting RPC Call. The Requester terminates the RPC
transaction and reports the result status to the RPC consumer. The
Requester's session implementation then determines the session ID and
slot for the failed request and performs slot recovery to make that
slot usable again. Otherwise, that slot is rendered permanently
unavailable.
When an NFS session is not present (for example, when NFS version 4.0
is in use), a transport error does not indicate whether the server
has processed the arguments of the RPC Call, or whether the server
has accessed or modified client memory associated with that RPC.
6. Upper-Layer Binding For NFS Version 4 Callbacks
The NFS version 4 family of protocols supports server-initiated
callbacks to notify NFS version 4 clients of events such as recalled
delegations.
6.1. NFS Version 4.0 Callback
An NFS version 4.0 client uses the SETCLIENTID operation for
advertising the IP address, port, and netid of its NFS version 4.0
callback service. When an NFS version 4.0 server provides a
backchannel service to an NFS version 4.0 client that uses RPC-over-
RDMA version 2 for its forward channel, the server MUST advertise the
backchannel service using either the "tcp" or "tcp6" netid.
Because the NFSv4.0 backchannel does not operate on RPC-over-RDMA,
this document does not specify an Upper-Layer binding for the NFSv4.0
backchannel RPC program.
6.2. NFS Version 4.1 Callback
In NFS version 4.1 and newer minor versions, callback operations may
appear on the same connection that is in use for NFS version 4
forward channel client requests. NFS version 4 clients and servers
MUST use the mechanisms described in Section 4.5 of
[I-D.ietf-nfsv4-rpcrdma-version-two] to convey backchannel operations
on an RPC-over-RDMA version 2 transport.
The csa_back_chan_attrs argument of the CREATE_SESSION operation
contains a ca_maxresponsesize field. The value in this field is the
absolute maximum size of backchannel replies generated by a replying
NFS version 4 client.
There are no DDP-eligible data items in callback procedures defined
in NFS version 4.1 or NFS version 4.2. However, some callback
operations, such as messages that convey device ID information, can
be sizeable. A sender can use Message Continuation or a Special
Payload message in this situation.
When an NFS version 4.1 client can support Special Payload Calls in
its backchannel, it reports a backchannel ca_maxrequestsize that is
larger than the connection's inline thresholds. Otherwise, an NFS
version 4 server MUST use only Simple Payload or Continued Payload
messages to convey backchannel operations.
7. Extending NFS Upper-Layer Bindings
RPC programs such as NFS must have an Upper-Layer Binding RPC programs such as NFS must have an Upper-Layer Binding
specification to operate on an RPC-over-RDMA version 2 transport specification to operate on an RPC-over-RDMA version 2 transport
[I-D.ietf-nfsv4-rpcrdma-version-two]. Via standards action, the [I-D.ietf-nfsv4-rpcrdma-version-two]. Via standards action, the
Upper-Layer Binding specified in this document can be extended to Upper-Layer Binding specified in this document can be extended to
cover versions of the NFS version 4 protocol specified after NFS cover versions of the NFS version 4 protocol specified after NFS
version 4 minor version 2, or to cover separately published version 4 minor version 2, or to cover separately published
extensions to an existing NFS version 4 minor version, as described extensions to an existing NFS version 4 minor version, as described
in [RFC8178]. in [RFC8178].
7. Security Considerations 8. Security Considerations
RPC-over-RDMA version 2 supports all RPC security models, including RPC-over-RDMA version 2 supports all RPC security models, including
RPCSEC_GSS security and transport-level security [RFC7861]. The RPCSEC_GSS security and transport-level security [RFC7861]. The
choice of what Direct Data Placement mechanism to convey RPC argument choice of what Direct Data Placement mechanism to convey RPC argument
and results does not affect this since it changes only the method of and results does not affect this since it changes only the method of
data transfer. Because the current document defines only the binding data transfer. Because the current document defines only the binding
of the NFS protocols atop RPC-over-RDMA version 2 of the NFS protocols atop RPC-over-RDMA version 2
[I-D.ietf-nfsv4-rpcrdma-version-two], all relevant security [I-D.ietf-nfsv4-rpcrdma-version-two], all relevant security
considerations are, therefore, described at that layer. considerations are, therefore, described at that layer.
8. IANA Considerations 9. IANA Considerations
The use of direct data placement in NFS introduces a need for an The use of direct data placement in NFS introduces a need for an
additional port number assignment for networks that share traditional additional port number assignment for networks that share traditional
UDP and TCP port spaces with RDMA services. The DDP protocol is such UDP and TCP port spaces with RDMA services. The DDP protocol is such
an example [RFC5041]. an example [RFC5041].
For this purpose, the current document lists a set of port number For this purpose, the current document lists a set of port number
assignments that IANA has already assigned for NFS/RDMA in the IANA assignments that IANA has already assigned for NFS/RDMA in the IANA
port registry, according to the guidelines described in [RFC6335]. port registry, according to the guidelines described in [RFC6335].
nfsrdma 20049/tcp Network File System (NFS) over RDMA nfsrdma 20049/tcp Network File System (NFS) over RDMA
nfsrdma 20049/udp Network File System (NFS) over RDMA nfsrdma 20049/udp Network File System (NFS) over RDMA
nfsrdma 20049/sctp Network File System (NFS) over RDMA nfsrdma 20049/sctp Network File System (NFS) over RDMA
The author requests that IANA add the current document as a reference The author requests that IANA add the current document as a reference
for the existing nfsrdma port assignments. This document does not for the existing nfsrdma port assignments. This document does not
alter these assignments. alter these assignments.
9. References 10. References
9.1. Normative References 10.1. Normative References
[I-D.ietf-nfsv4-rpcrdma-version-two] [I-D.ietf-nfsv4-rpcrdma-version-two]
Lever, C. and D. Noveck, "RPC-over-RDMA Version 2 Lever, C. and D. Noveck, "RPC-over-RDMA Version 2
Protocol", Work in Progress, Internet-Draft, draft-ietf- Protocol", Work in Progress, Internet-Draft, draft-ietf-
nfsv4-rpcrdma-version-two-05, 6 July 2021, nfsv4-rpcrdma-version-two-06, 2 January 2022,
<https://datatracker.ietf.org/doc/html/draft-ietf-nfsv4- <https://datatracker.ietf.org/doc/html/draft-ietf-nfsv4-
rpcrdma-version-two-05>. rpcrdma-version-two-06>.
[RFC1833] Srinivasan, R., "Binding Protocols for ONC RPC Version 2", [RFC1833] Srinivasan, R., "Binding Protocols for ONC RPC Version 2",
RFC 1833, DOI 10.17487/RFC1833, August 1995, RFC 1833, DOI 10.17487/RFC1833, August 1995,
<https://www.rfc-editor.org/rfc/rfc1833>. <https://www.rfc-editor.org/rfc/rfc1833>.
[RFC2119] Bradner, S., "Key words for use in RFCs to Indicate [RFC2119] Bradner, S., "Key words for use in RFCs to Indicate
Requirement Levels", BCP 14, RFC 2119, Requirement Levels", BCP 14, RFC 2119,
DOI 10.17487/RFC2119, March 1997, DOI 10.17487/RFC2119, March 1997,
<https://www.rfc-editor.org/rfc/rfc2119>. <https://www.rfc-editor.org/rfc/rfc2119>.
skipping to change at page 16, line 45 skipping to change at page 17, line 26
[RFC8174] Leiba, B., "Ambiguity of Uppercase vs Lowercase in RFC [RFC8174] Leiba, B., "Ambiguity of Uppercase vs Lowercase in RFC
2119 Key Words", BCP 14, RFC 8174, DOI 10.17487/RFC8174, 2119 Key Words", BCP 14, RFC 8174, DOI 10.17487/RFC8174,
May 2017, <https://www.rfc-editor.org/rfc/rfc8174>. May 2017, <https://www.rfc-editor.org/rfc/rfc8174>.
[RFC8881] Noveck, D., Ed. and C. Lever, "Network File System (NFS) [RFC8881] Noveck, D., Ed. and C. Lever, "Network File System (NFS)
Version 4 Minor Version 1 Protocol", RFC 8881, Version 4 Minor Version 1 Protocol", RFC 8881,
DOI 10.17487/RFC8881, August 2020, DOI 10.17487/RFC8881, August 2020,
<https://www.rfc-editor.org/rfc/rfc8881>. <https://www.rfc-editor.org/rfc/rfc8881>.
9.2. Informative References 10.2. Informative References
[RFC1094] Nowicki, B., "NFS: Network File System Protocol [RFC1094] Nowicki, B., "NFS: Network File System Protocol
specification", RFC 1094, DOI 10.17487/RFC1094, March specification", RFC 1094, DOI 10.17487/RFC1094, March
1989, <https://www.rfc-editor.org/rfc/rfc1094>. 1989, <https://www.rfc-editor.org/rfc/rfc1094>.
[RFC1813] Callaghan, B., Pawlowski, B., and P. Staubach, "NFS [RFC1813] Callaghan, B., Pawlowski, B., and P. Staubach, "NFS
Version 3 Protocol Specification", RFC 1813, Version 3 Protocol Specification", RFC 1813,
DOI 10.17487/RFC1813, June 1995, DOI 10.17487/RFC1813, June 1995,
<https://www.rfc-editor.org/rfc/rfc1813>. <https://www.rfc-editor.org/rfc/rfc1813>.
skipping to change at page 17, line 24 skipping to change at page 18, line 7
[RFC8178] Noveck, D., "Rules for NFSv4 Extensions and Minor [RFC8178] Noveck, D., "Rules for NFSv4 Extensions and Minor
Versions", RFC 8178, DOI 10.17487/RFC8178, July 2017, Versions", RFC 8178, DOI 10.17487/RFC8178, July 2017,
<https://www.rfc-editor.org/rfc/rfc8178>. <https://www.rfc-editor.org/rfc/rfc8178>.
[XNFS] The Open Group, "Protocols for Interworking: XNFS, Version [XNFS] The Open Group, "Protocols for Interworking: XNFS, Version
3W", January 1998. 3W", January 1998.
Acknowledgments Acknowledgments
Thanks to Tom Talpey, who contributed the text of Section 5.4.2. Thanks to Tom Talpey, who contributed the text of Section 5.1.2.2.
David Noveck contributed the text of Section 5.6 and Section 6. The David Noveck contributed the text of Section 5.5 and Section 7. The
author also wishes to thank Bill Baker and Greg Marsden for their author also wishes to thank Bill Baker and Greg Marsden for their
support of this work. support of this work.
Special thanks go to Transport Area Directors Zaheduzzaman Sarker, Special thanks go to Transport Area Directors Zaheduzzaman Sarker,
NFSV4 Working Group Chairs Brian Pawlowski, and David Noveck, and NFSV4 Working Group Chairs Brian Pawlowski, and David Noveck, and
NFSV4 Working Group Secretary Thomas Haynes for their support. NFSV4 Working Group Secretary Thomas Haynes for their support.
Author's Address Author's Address
Charles Lever Charles Lever
 End of changes. 46 change blocks. 
180 lines changed or deleted 194 lines changed or added

This html diff was produced by rfcdiff 1.48. The latest version is available from http://tools.ietf.org/tools/rfcdiff/