NFSv4 Working Group S. Faibish
Internet-Draft EMC Corporation
Intended status: draft D. Black
Expires: September 3 2010 EMC Corporation
M. Eisler
NetApp
J. Glasgow
Google
March 5, 2010
pNFS Access Permissions Check
draft-faibish-nfsv4-pnfs-access-permissions-check-02
Status of this Memo
This Internet-Draft is submitted to IETF in full conformance with the
provisions of BCP 78 and BCP 79.
Internet-Drafts are working documents of the Internet Engineering
Task Force (IETF), its areas, and its working groups. Note that
other groups may also distribute working documents as Internet-
Drafts.
Internet-Drafts are draft documents valid for a maximum of six months
and may be updated, replaced, or obsoleted by other documents at any
time. It is inappropriate to use Internet-Drafts as reference
material or to cite them other than as "work in progress."
The list of current Internet-Drafts can be accessed at
http://www.ietf.org/ietf/1id-abstracts.txt
The list of Internet-Draft Shadow Directories can be accessed at
http://www.ietf.org/shadow.html
This Internet-Draft will expire on September 5, 2010.
Copyright Notice
Copyright (c) 2010 IETF Trust and the persons identified as the
document authors. All rights reserved.
This document is subject to BCP 78 and the IETF Trust's Legal
Provisions Relating to IETF Documents (http://trustee.ietf.org/license-
info) in effect on the date of publication of this document.
Please review these documents carefully, as they describe your rights
and restrictions with respect to this document.
Faibish et al. Expires September 5, 2010 [Page 1]
Internet-Draft pNFS Access Permissions Check March 2010
Abstract
This document extends the pNFS protocol to communicate the results of
permission checks for access to the data servers referenced by
layouts, including checks performed by both clients and the MDS. The
extension provides means for clients to communicate client-detected
access denial errors to the MDS, including the case in which a client
requests direct NFS access via the MDS that the MDS cannot perform.
Table of Contents
1. Introduction...................................................3
1.1. Example...................................................4
1.2. Issues with the current pNFS protocol.....................5
1.2.1. Client access permission denial to SD................5
1.2.2. MDS access permission denial to SD...................6
1.2.3. New MDS Requirement..................................6
2. Conventions used in this document..............................6
3. Description of the proposed approaches to solution.............6
3.1. Defining the opaque fields of LAYOUTRETURN................7
3.1.1. ARGUMENT.............................................7
3.1.2. RESULT...............................................8
3.1.3. Description..........................................8
3.2. Implementation using a new layoutreturn_type4.............9
3.2.1. ARGUMENT.............................................9
3.2.2. RESULT...............................................9
3.2.3. New LAYOUTRETURN type description...................10
3.3. Operation: CB_LAYOUTACCESSCHECKRECALL - Ask client to check
permissions...................................................10
3.3.1. ARGUMENT............................................10
3.3.2. RESULT..............................................11
3.3.3. DESCRIPTION.........................................11
4. Reporting storage device inaccessibility......................11
4.1. Access denied to client at mount time....................11
4.2. Permission denied to the client at I/O time..............12
4.2.1. pNFS client detects permission access denial........13
4.2.2. Layout command that require permissions check by the
client.....................................................13
4.2.2.1. Case 1 - MDS successfully performs I/O to the
device..................................................13
4.2.2.2. Case 2 - MDS fails to perform the I/O to the device
........................................................14
4.3. Permission denied to MDS server at I/O time..............14
5. Security Considerations.......................................14
6. IANA Considerations...........................................15
Faibish et al. Expires September 5, 2010 [Page 2]
Internet-Draft pNFS Access Permissions Check March 2010
7. Conclusions...................................................15
8. References....................................................15
8.1. Normative References.....................................15
8.2. Informative References...................................15
9. Acknowledgments...............................................16
Authors' Addresses...............................................17
1. Introduction
Figure 1 shows the overall architecture of a Parallel NFS (pNFS)
system:
+-----------+
|+-----------+ +-----------+
||+-----------+ | |
||| | NFSv4.1 + pNFS | |
+|| Clients |<------------------------------>| MDS |
+| | | |
+-----------+ | |
||| +-----------+
||| |
||| |
||| Storage +-----------+ |
||| Protocol |+-----------+ |
||+----------------||+-----------+ Control |
|+-----------------||| | Protocol |
+------------------+|| Storage |------------+
+| Devices |
+-----------+
Figure 1 pNFS Architecture
Inconsistent access permissions expose a gap in the pNFS protocol.
The pNFS protocol assumes that a client can access every storage
device (SD) included in a valid layout sent by the MDS server, and
provides no means to communicate client access failure to the MDS. It
has been argued that this is an implementation detail, but access
failures permission denials can impair the performance scalability
value of pNFS and allow errors to go unreported. There is no pNFS
error mechanism to inform a system administrator that a client lacks
permission to access a storage device at mount time or when I/O is
performed. There is a related problem when an MDS doesn't have access
permission to some storage devices and hence cannot perform I/O on
behalf of a client. In this document storage devices are a generic
term for data servers and/or storage servers used by the file, block
and object pNFS layouts.
Faibish et al. Expires September 5, 2010 [Page 3]
Internet-Draft pNFS Access Permissions Check March 2010
In the case of the block layout [RFC5663] if the MDS has no access to
a storage device (LUN) implementations are generally unable to export
the NFS mount point for any filesystem using that storage device. In
this situation, clients will be unable to mount that file system and
an error will presumably be logged by the MDS server. If the MDS can
access all the storage devices involved, but the client doesn't have
sufficient access to some storage devices/LUNs, at mount time the
client may choose to mount the file system using NFSV4.1 without pNFS
support (fallback to NFS). This failure to mount as a pNFS file
system cannot currently be communicated to the server because there
are no protocol messages defined which convey this failure.
The above discussion also applies to the file and object layout pNFS
clients regardless of whether the MDS has permissions to access the
storage devices, with one important difference. In contrast to the
block layout, MDSs for the file and object layouts are often unable
to access the storage devices that store data for the exported
filesystem. This make it significantly easier for a file or object
layout MDS to provide layouts that contain inaccessible devices, in
contrast to the block layout where an MDS should not allow a client
to mount a FSID to which the MDS has no access permission.
There is no error reporting mechanism in the pNFS protocol for this
type of error. Even if we correct the access permission issue the
introduction of a new error reporting mechanism at I/O time for both
server and client can be problematic as it may be too chatty. We
propose to introduce a new error case but leave the error reporting
mechanism at I/O time OPTIONAL or an optimization to the latitude of
the server and client implementation.
Although the change to the protocol is delicate, logging some kind of
warning at the client might be appropriate as an implementation
option on the client to reduce chattiness.
1.1. Example
A motivating use case is addition of a new storage device to which
all the pNFS clients (1000s of them) lack access permission. Layouts
cannot be granted that use this new device, requiring that all I/Os
to that new storage device be served by the MDS server creating a
performance and scalability bottleneck that may be difficult to
detect based on I/O behavior.
A better approach to this issue is to report the access failure
before the client attempts to issue any I/Os that can only be
Faibish et al. Expires September 5, 2010 [Page 4]
Internet-Draft pNFS Access Permissions Check March 2010
serviced by the MDS server. This makes the problem explicit, rather
than the forcing the MDS, or a system administrator to diagnose the
performance problem caused by client I/O using NFS instead of the
pNFS layout. There are limits to this approach because complex mount
structures may prevent a client from detecting this situation at
mount time, but at a minimum, access problems involving the root of
the mount structure can be detected. See section 1.2.1 for a detailed
example.
This document adds error reporting mechanisms to address both this
situation and situations in which the client cannot detect the access
problem until it attempts to perform I/O to the inaccessible storage
device.
1.2. Issues with the current pNFS protocol
Scenario of Interest: Client expects to be able to use pNFS (e.g.,
use -pnfs switch to mount command, or similar), but one or more
storage devices are inaccessible. This discussion does not apply to
a client that doesn't care whether pNFS is used (e.g., uses pNFS to
optimize if available, but for which it is acceptable that access is
performed via the main NFS server).
Desired client behavior: Client gets the entire storage device list
for a mount point from server and checks it as part of the mount
operation (or at whatever point it first realizes that it expects to
use pNFS).
Missing protocol functionality: Client has no obvious way to report
an inaccessible storage device to the server.
1.2.1. Client access permission denial to SD
A client doesn't communicate to the MDS server that the client's
access to a storage device is denied as a result of an access
permission issue. When the pNFS server grants a layout to the client,
it assumes the client can access the storage devices (files, LUNs, or
objects). The server cannot check this because the server cannot
issue I/Os via the client and because connectivity is not transitive
- the client may have good network connectivity to the MDS, the MDS
may have good storage connectivity to the storage devices, but
something prevents the client from talking to one or more of the
storage devices. This could be a network mis-configuration or
failure, and is a possible scenario for all pNFS layout types.
This access permission problem cannot be reported by the MDS server
when the client mounts the filesystem for several reasons. First, the
Faibish et al. Expires September 5, 2010 [Page 5]
Internet-Draft pNFS Access Permissions Check March 2010
MDS pNFS server doesn't know whether the client supports or intends
to use pNFS. Second, the MDS NFS server doesn't know that the client
is mounting the NFS filesystem (there is no explicit mount operation
in NFSv4). Third, it is unreasonable to expect the MDS to know and
check the entire mount structure below the mount point used by the
client. For example, if the client mounts "/", the file systems below
"/" may have pNFS capabilities, but refer to different storage
devices. Or the client may mount say "/a/b/c/d", where "d" uses a
pNFS capable storage device, but the client subsequently does I/O to
"e/f/g/h/i/j/k", where "k" is either not pNFS capable or uses a
storage device different from the storage device used by "d".
1.2.2. MDS access permission denial to SD
The current pNFS server protocol doesn't require MDS data access to
the storage devices. Although the MDS is not required to check
permissions, it is assumed that the devices are correctly configured
when the pNFS filesystem is initialized on the MDS server and
exported. Even if the administrator checks the MDS access permission
to all storage devices during initial configuration, the problem may
surface at a later point in time when a new storage device is added
or other changes are made. For the specific case of adding a new
storage device, an MDS check of I/Os to the newly added device before
using it in layouts avoids this set of problems, but this does not
cover loss of MDS access to existing storage devices.
1.2.3. New MDS Requirement
The metadata server (MDS) SHOULD NOT use storage devices in pNFS
layouts that are not accessible to the MDS. To the extent that an
MDS can determine whether storage devices are accessible to clients,
if a client cannot access a storage device, an MDS SHOULD NOT include
that storage device in a pNFS layouts sent to that client.
2. Conventions used in this document
The key words "MUST", "MUST NOT", "REQUIRED", "SHALL", "SHALL NOT",
"SHOULD", "SHOULD NOT", "RECOMMENDED", "MAY", and "OPTIONAL" in this
document are to be interpreted as described in RFC-2119 [RFC2119].
3. Description of the proposed approaches to solution
There are several possible solutions. The first is to implement a
new operation, LAYOUTRETURN4x that returns layouts to the MDS along
with error information. Clients that receive an NFS4ERR_NOTSUPP
error SHOULD mark the server as not supporting this operation and use
LAYOUTRETURN instead.
Faibish et al. Expires September 5, 2010 [Page 6]
Internet-Draft pNFS Access Permissions Check March 2010
Another possible approach is to make use of the opaque field
available in LAYOUTRETURN. One could define part of this field for
all layout types. In the case that the pNFS client has a valid
layout on a file but cannot perform I/O to a SD due to lack of access
permission, the client will fall back the I/O to the MDS NFS server.
Before the client sends the I/O to the NFS server it sends a
LAYOUTRETURN command for the purpose of avoiding unnecessary MDS
CB_LAYOUTRECALL operations in the future. The client sends the
LAYOUTRETURN operation for every layouts that uses to the
inaccessible SD and includes an error reporting that the reason for
the fall back to the NFS server is an access permission denial to the
specific deviceid4. The client may return disjoint regions of the
file by using multiple LAYOUTRETURN operations within a single
COMPOUND operation. The client will include NFS4ERR_DEVICE_PERM_DENY
in the new LAYOUTRETURN operation.
A third approach is to introduce a new LAYOUTRETURN type at FSID
scope such as LAYOUT4_RET_REC_FSID_NO_ACCESS, i.e., return all
layouts for this FSID and tell the server that the reason for the
return is a connectivity issue. In order to differentiate the
permission issue from a real connectivity issue the solution will
require the client to do two LAYOUTRETURN operations to deal with
servers that don't understand the new type. The two LAYOUTRETURN
operations happen once per client using
LAYOUT4_RET_REC_FSID_NO_ACCESS and only in an error case followed by
a second operation for the existing FSID scope to interoperate with
an MDS that doesn't understand the new scope.
3.1. Defining the opaque fields of LAYOUTRETURN
3.1.1. ARGUMENT
When the LAYOUTRETURN operation specifies a LAYOUTRETURN4_FILE_return
type, then the layoutreturn_file4 data structure specifies the region
of the file layout that is no longer needed by the client. For each
layout type we define the opaque lrf_body so that it can communicate
an error code to the server as well as the deviceid4 which
encountered the error. This has already been defined for the object
layout type [RFC5664]
For the file layout we define the opaque body as follows:
struct nfsv4_1_file_layoutreturn4 {
deviceid4 lrf_deviceid;
nfsstat4 lrf_status;
};
Faibish et al. Expires September 5, 2010 [Page 7]
Internet-Draft pNFS Access Permissions Check March 2010
An MDS server should check the length of the lrf_body. If the length
is zero, then the client has not communicated additional information
with the layout return. This will generally be the case when a file
is closed, or in response to a CB_LAYOUTRECALL operation.
For the block layout type, we similarly define the block specific
structure as:
struct pnfs_block_layoutreturn4 {
deviceid4 lrf_deviceid;
nfsstat4 lrf_status;
};
The alternative, which is more complex is to make the status (error)
and deviceid4 common to all LAYOUTRETURN operations, but do so by
adding a new operation or a new return
type(LAYOUT4_RET_REC_FILE_ERROR)
struct layoutreturn_file_error4 {
offset4 lrf_offset;
length4 lrf_length;
stateid4 lrf_stateid;
deviceid4 lrf_deviceid;
nfsstat4 lrf_status;
/* layouttype4 specific data */
opaque lrf_body<>;
};
3.1.2. RESULT
The LAYOUTRETURN4res remains unchanged.
3.1.3. Description
This solution will add a new error case to LAYOUTRETURN. The
implementation will use LAYOUTRETURN when FSID is sent to the client.
When the client fails an I/O as a result of access permission denial
it will send a LAYOUTRETURN operation to the MDS server with new
error NFS4ERR_DEVICE_PERM_DENY and the deviceid4 on which access
permission was denied.
When the server receives this error it MAY log an error to the syslog
and perform an access permission check to the SD expecting that the
Faibish et al. Expires September 5, 2010 [Page 8]
Internet-Draft pNFS Access Permissions Check March 2010
client will fall back the I/O to the MDS. If the permission check of
the server fails the NFS4ERR_DEVICE_PERM_DENY SHOULD be logged.
3.2. Implementation using a new layoutreturn_type4
In this section we will define the use case addressed by this
implementation.
3.2.1. ARGUMENT
/* Constants used for new LAYOUTRETURN and CB_LAYOUTRECALL */
const LAYOUT4_RET_REC_FILE = 1;
const LAYOUT4_RET_REC_FSID = 2;
const LAYOUT4_RET_REC_ALL = 3;
const LAYOUT4_RET_REC_FSID_NO_ACCESS = 4;
enum layoutreturn_type4 {
LAYOUTRETURN4_DEVICE = LAYOUT4_RET_REC_FSID_NO_ACCESS,
LAYOUTRETURN4_FILE = LAYOUT4_RET_REC_FILE,
LAYOUTRETURN4_FSID = LAYOUT4_RET_REC_FSID,
LAYOUTRETURN4_ALL = LAYOUT4_RET_REC_ALL
};
struct layoutreturn_device4 {
offset4 lrf_offset;
length4 lrf_length;
stateid4 lrf_stateid;
deviceid4 lrf_deviceid;
nfsstat4 lrf_status;
/* layouttype4 specific data */
opaque lrf_body<>;
};
union layoutreturn4 switch(layoutreturn_type4 lr_returntype) {
case LAYOUTRETURN4_DEVICE:
layoutreturn_device4 lr_layout;
default: void;
};
3.2.2. RESULT
union LAYOUTRETURN4res switch (nfsstat4 lorr_status) {
Faibish et al. Expires September 5, 2010 [Page 9]
Internet-Draft pNFS Access Permissions Check March 2010
case NFS4_OK:
layoutreturn_stateid lorr_stateid;
default:
void;
};
3.2.3. New LAYOUTRETURN type description
We will use a new LAYOUTRETURN layoutreturn_type4, let's call it
LAYOUT4_RET_REC_FSID_NO_ACCESS, in which case the client returns all
layouts for this FSID and informs the server that the reason for the
return is an inability to access the device. The same stateid may be
used or in order to report a new error client will force a new
stateid. We will also add the mechanism to report a new error
NFS4ERR_DEVICE_PERM_DENY.
Backwards compatibility may require a client to do two layout return
operations to deal with servers that don't understand the new
layoutreturn_type4. If the server doesn't understand the new
layoutreturn_type4, then the server will respond with an error code.
The client SHOULD do an ordinary FSID return and remember that the
new return type is not to be used with this server. This assumes that
the client is sufficiently disrupted by the problem to decide to drop
all layouts for the filesystem (FSID). Alternatively, for servers
that understand the new layoutreturn when the server receives a new
stateid it will check if there is an NFS4ERR_DEVICE_PERM_DENY error
or issue an CB_LAYOUTRECALL to get the error code from the client.
3.3. Operation: CB_LAYOUTACCESSCHECKRECALL - Ask client to check
permissions
3.3.1. ARGUMENT
/*
* NFSv4.1 callback arguments and results
/*
struct CB_LAYOUTACCESSCHECK4args {
nfs_fh4 claca_fh;
offset4 claca_offsets[];
};
Faibish et al. Expires September 5, 2010 [Page 10]
Internet-Draft pNFS Access Permissions Check March 2010
3.3.2. RESULT
struct layoutaccesscheck_device4 {
deviceid4 lac_device_id;
nfsstat4 lac_status;
};
struct CB_LAYOUTACCESSCHECK4res {
layoutaccesscheck_device4 clacr_status[];
};
3.3.3. DESCRIPTION
In this case the client checks that it has permission access to all
the deviceid that are included in all the layouts in his possession
and report to the MDS deviceid with permission access denial. Using
this operation the MDS will find out what are the SDs that have
permission access issues for more than one client that have valid
layouts to that device and didn't yet found that there is a
permission access issue. In this case the MDS can prevent the client
from falling back to NFS by recalling the layout and removing the
faulty device from the layout thus preventing a storm of I/Os to the
MDS. The MDS will only send a CB_LAYOUTACCESSCHECK command to clients
that already have a valid layout for the faulty device. As an
implementation recommendation the MDS will remove that device from
the valid devices list and will log an error mentioning that there is
a problem with that device. All the layouts delivered to new client
requests will exclude the device with the problem. Some servers may
chose to perform the I/O via the MDS with the risk of a retry and I/O
error of the MDS. In this latest case the MDS will unilaterally
remove that device from the list and will recall all the layouts from
all the clients that have layouts to that device and send new layouts
excluding the faulty device.
4. Reporting storage device inaccessibility
4.1. Access denied to client at mount time
The most suitable time for the client reporting access denial by a
data server is at the mount time. This would be the preferred way to
address the issue but it is not possible with the current protocol
for several reasons: If the server initiates the request, MDS doesn't
know if the client wants to use pNFS or NFS. If the client is the
initiator of the error the client is mounting the pNFS filesystem
Faibish et al. Expires September 5, 2010 [Page 11]
Internet-Draft pNFS Access Permissions Check March 2010
knowing that it will use pNFS for access but the client doesn't
specifically request pNFS.
The solution requires a special tag -pnfs or a switch to the mount
command and syscall at the client. The client cannot explicitly
request pNFS as it needs first to discover that the server is
supporting pNFS by sending a pNFS LAYOUTGET request to the server at
mount time. If this request fails, the resulting error cannot be
reported by the client.
The client will send an OPEN request to access a file and to find if
the NFS server is pNFS capable it will send a LAYOUTGET command and
if the NFS server doesn't accept and returns an error the client will
request access using plain NFS. The client will decide if this is an
error or not. In the case that the LAYOUTGET command succeeded the
client may still ask the MDS to deliver the I/O. So, inherently the
client has to query the MDS for access permissions to all the SDs
that are used in the layout sent to the client before accessing the
deviceid included in the layout. The pNFS protocol doesn't require
the MDS to check access permission to the devices that are included
in the layout. It is assumed that the MDS has permission access to
all the devices it includes in the layout without any checks.
If the MDS doesn't know if it has access or not to a deviceid it
shouldn't put that device in the layout granted to clients in order
to prevent cases when the client sends the I/O using plain NFS from
the MDS. If the MDS doesn't have permission access to a SD it will
send an error to the client and the I/O will fail. Based on the above
behavior the best time to check is at the time when the initial
configuration of the pNFS filesystem is done. Currently the pNFS spec
states that a client can write through or read from the MDS, whether
it has a layout or not or it does not support pNFS assuming that the
MDS has permission access to all the SDs. We propose to make this
implicit recommendation explicit.
4.2. Permission denied to the client at I/O time
In this case when the pNFS capable client receives a valid layout
from the MDS server and cannot write to the storage devices, the
client falls back to the NFS server to perform the I/O. There is no
error currently logged by the client or sent back to the MDS server
in this situation. The client will use the new error case added in
section 3.1 or will use a LAYOUTRETURN including
NFS4ERR_DEVICE_PERM_DENY error code as defined in section 3.2. If the
client didn't access the SD that has the permission denial yet and it
is not aware of such an issue the client couldn't send an error to
the MDS. But if the MDS got a permission error for a deviceid from
Faibish et al. Expires September 5, 2010 [Page 12]
Internet-Draft pNFS Access Permissions Check March 2010
another client it can send a CB_LAYOUTRECALL with FSID_PERM_CHECK to
the client in the case when a pNFS client requests the MDS to write
an I/O to one of the devices from a layout sent to the client by the
MDS before. The client will send a LAYOUTRETURN and the MDS will
check that the error is NFS4ERR_DEVICE_PERM_DENY and to confirm that
this is a permission access issue not a connectivity or other error.
4.2.1. pNFS client detects permission access denial
The current protocol recommends that the client fallback to the
NFSv4.1 server to do the I/O on the behalf of the client and in same
compound command it includes a LAYOUTRETURN command for the layout
part on the Storage Device with permission issues. The recommendation
can be interpreted as LAYOUTRETURN of all the layouts for that file.
According to the section 3.2 in this case the client will issue a
LAYOUTRETURN mandatory for the layout offset in the range that
resides on the deviceid with permission access denial together with
the fallback of I/O to NFS. An error code NFS4ERR_DEVICE_PERM_DENY
will be included in the LAYOUTRETUN command. In general fallback to
NFS is restricted to the cases of server or client failure recovery.
In this case the fallback will be related to the permission access
issue as an additional case of fallback to NFS.
4.2.2. Layout command that require permissions check by the client
Assume there is a list of devices used by a given file. The client
attempts a write operation and fails with a permission error. The
client will retry (fallback) the I/O via the metadata server.
For block layout type, the client SHOULD return the layout before
attempting to retry the I/O via the MDS. Object and file clients,
need not return the layout before attempting to retry the I/O via the
MDS.
If the client returns the layout, the client SHOULD indicate which
device caused an error (or the range of the file in which the error
occurred).
4.2.2.1. Case 1 - MDS successfully performs I/O to the device
MDS proactively sends an CB_LAYOUT_ACCESS_CHECK to all clients that
have a layout referencing the storage device which recently returned
a permission access error. The CB _LAYOUT_ACCESS_CHECK will contain a
file handle, and a list of offsets. For file layout, the client can
compute the data servers to which it must send an NFS ACCESS command.
The client SHOULD issue the NFS ACCESS command on behalf of any one
of the users that have the file currently open on each client. The
Faibish et al. Expires September 5, 2010 [Page 13]
Internet-Draft pNFS Access Permissions Check March 2010
client should then accumulate the results of all the access checks
(there may be more than one device checked). The client returns a
vector of device handles and statuses.
The STATUS code is either NFS4_OK or the error code returned by the
data store. The implementation details of how the server aggregates
the client responses to CB_LAYOUT_ACCESS_CHECK is left as an exercise
for the reader. In many instances if the server detects that a
majority, or a large number of clients cannot access some devices,
the server will issue CB_LAYOUT_RECALL to all the clients, if
possible it will restripe (or re-layout) the file to exclude the
failing device.
4.2.2.2. Case 2 - MDS fails to perform the I/O to the device
This is the same as case 1, except that the server can restripe the
file, only if the failed device does not yet contain data for the
file. Implementations may decide to remove the failing device from
the list of devices used for new files.
4.3. Permission denied to MDS server at I/O time
In case when the client holding a valid layout requests the NFS
server to execute the I/O and the MDS gets an access permission
denial from the storage device, the MDS cannot perform the I/O and
returns an error to the client. In this case all client I/O to that
device will fail and the reason for these failures needs to be
communicated to the MDS. To address this case the client will use the
new layoutreturn_type4 operation defined in section 3.2 and the new
NFS4ERR_DEVICE_PERM_DENY error code to inform the MDS of possible
permission access issues. An additional option is to use the new
CB_LAYOUTACCESSCHECKRECALL from section 3.3sent to the client
requesting permission access check. On failure the MDS will log an
error NFS4ERR_DEVICE_UNACCESSIBLE to inform the admin to correct the
problem. On receiving the CB the client will send the SD a
GETDEVICEINFO and report NFS4ERR_DEVICE_PERM_DENY to the MDS server
using the new layout command from section 3.2.
5. Security Considerations
All control operations from the MDS to the storage devices, including
any operations required for access permission checks in order to
detect permission denials to the MDS and the pNFS client, SHOULD be
authenticated in order to address cases when the access permission is
denied to the client by the administrator. It is expected that the
permission denial to a certain data server to a certain client will
Faibish et al. Expires September 5, 2010 [Page 14]
Internet-Draft pNFS Access Permissions Check March 2010
be known to the MDS by configuration. This is applicable to all pNFS
layout types.
6. IANA Considerations
There are no IANA considerations in this document beyond pNFS IANA
Considerations are covered in [RFC5661].
7. Conclusions
This draft specifies additions to the pNFS protocol addressing access
permission checks of the client and MDS server to storage devices
used in pNFS layouts for all layout types.
8. References
8.1. Normative References
[LEGAL] IETF Trust, "Legal Provisions Relating to IETF
Documents",URL http://trustee.ietf.org/docs/IETF-Trust-
License-Policy.pdf, November 2008.
[RFC2119] Bradner, S., "Key words for use in RFCs to Indicate
Requirement Levels", BCP 14, RFC 2119, March 1997.
[RFC5661] Shepler, S., Eisler, M., and D. Noveck, "Network File
System (NFS) Version 4 Minor Version 1 Protocol",
http://tools.ietf.org/html/rfc5661, January 2010.
[RFC5663] Black, D., Glasgow, J., Fridella, S., "Parallel NFS (pNFS)
Block/Volume Layout", http://tools.ietf.org/html/rfc5663,
January 2010.
[RFC5664] Halevy, B., Welch, B., Zelenka, J., "Object-Based Parallel
NFS (pNFS) Operations", http://tools.ietf.org/html/rfc5664,
January 2010
[XDR] Eisler, M., "XDR: External Data Representation Standard",
STD 67, RFC 4506, May 2006.
8.2. Informative References
[MPFS] EMC Corporation, "EMC Celerra Multi-Path File System", EMC
Data Sheet, available at:
http://www.emc.com/collateral/software/data-sheet/h2006-
celerra-mpfs-mpfsi.pdf
link checked 16 October 2009
Faibish et al. Expires September 5, 2010 [Page 15]
Internet-Draft pNFS Access Permissions Check March 2010
9. Acknowledgments
This draft includes ideas from discussions with the authors of the
different pNFS layouts Jason Glasgow and Benny Halevy as well as pNFS
maintainer of Linux kernel including Bruce Fields.
This document was prepared using 2-Word-v2.0.template.dot.
Faibish et al. Expires September 5, 2010 [Page 16]
Internet-Draft pNFS Access Permissions Check March 2010
Authors' Addresses
Sorin Faibish (editor)
EMC Corporation
32 Coslin Drive
Southboro, MA 01772
US
Phone: +1 (508) 305-8545
Email: sfaibish@emc.com
David L. Black
EMC Corporation
176 South Street
Hopkinton, MA 01748
US
Phone: +1 (508) 293-7953
Email: black_david@emc.com
Michael Eisler
NetApp
5765 Chase Point Circle
Colorado Springs, CO 80919
US
Phone: +1 (719) 599 8759
Email: mike@eisler.com
Jason Glasgow
Google
5 Cambridge Center, Floors 3-6
Cambridge, MA 02142
US
Phone: +1 (617) 575 1300
Email: jglasgow@google.com
Faibish et al. Expires September 5, 2010 [Page 17]