They're in use in downstream for Ceph, and have been tested by QA, so
they should be safe. If we decide we don't want them in 2.5, then
downstream RHCS will just need to carry them as patches.
Daniel
On 11/02/2017 03:56 AM, Malahal Naineni wrote:
Dan, I remember that we waited for the recovery code (aka IP failover
code) reorganization patches to go into V2.6 alone. Do they now have
enough runtime to get merged into V2.5 stable branch?
Regards, Malahal.
On Tue, Oct 31, 2017 at 11:37 PM, Daniel Gryniewicz <d...@redhat.com
<mailto:d...@redhat.com>> wrote:
Here's the set of commits that downstream Ceph needs. Gluster can
also use the non-Ceph related ones.
Note, these are oldest first, not newest first.
Daniel
commit b862fe360b2a0f1b1d9d5d6a8b91f1550b66b269
Author: Gui Hecheng <guihech...@cmss.chinamobile.com
<mailto:guihech...@cmss.chinamobile.com>>
AuthorDate: Thu Mar 30 10:44:25 2017 +0800
Commit: Frank S. Filz <ffilz...@mindspring.com
<mailto:ffilz...@mindspring.com>>
CommitDate: Fri Aug 11 14:31:22 2017 -0700
SAL: extract fs logic from nfs4_recovery
This is a prepare patch for modulized recovery backends.
- define recovery apis: struct nfs_recovery_backend
- define hooks for recovery_fs module
Change-Id: I45523ef9a0e6f9a801fc733b095ba2965dd8751b
Signed-off-by: Gui Hecheng <guihech...@cmss.chinamobile.com
<mailto:guihech...@cmss.chinamobile.com>>
commit cb787a1cf4a4df4da672c6b00cb0724db5d99e4d
Author: Gui Hecheng <guihech...@cmss.chinamobile.com
<mailto:guihech...@cmss.chinamobile.com>>
AuthorDate: Thu Mar 30 10:50:18 2017 +0800
Commit: Frank S. Filz <ffilz...@mindspring.com
<mailto:ffilz...@mindspring.com>>
CommitDate: Fri Aug 11 14:31:23 2017 -0700
SAL: introduce new recovery backend based on rados kv store
Use rados OMAP API to implement a kv store for client tracking data
Change-Id: I1aec1e110a2fba87ae39a1439818a363b6cfc822
Signed-off-by: Gui Hecheng <guihech...@cmss.chinamobile.com
<mailto:guihech...@cmss.chinamobile.com>>
commit fbc905015d01a7f2548b81d84f35b76524543f13
Author: Gui Hecheng <guihech...@cmss.chinamobile.com
<mailto:guihech...@cmss.chinamobile.com>>
AuthorDate: Wed May 3 09:58:34 2017 +0800
Commit: Frank S. Filz <ffilz...@mindspring.com
<mailto:ffilz...@mindspring.com>>
CommitDate: Fri Aug 11 14:31:23 2017 -0700
cmake: make modulized recovery backends compile as modules
- add USE_RADOS_RECOV option for new rados kv backend
- keep original fs backend as default
Change-Id: I26c2c4f9a433e6cd70f113fa05194d6817b9377a
Signed-off-by: Gui Hecheng <guihech...@cmss.chinamobile.com
<mailto:guihech...@cmss.chinamobile.com>>
commit eb4eea1343251f17fe39de48426bc4363eaef957
Author: Gui Hecheng <guihech...@cmss.chinamobile.com
<mailto:guihech...@cmss.chinamobile.com>>
AuthorDate: Thu May 4 22:43:17 2017 +0800
Commit: Frank S. Filz <ffilz...@mindspring.com
<mailto:ffilz...@mindspring.com>>
CommitDate: Fri Aug 11 14:31:23 2017 -0700
config: add new config options for rados_kv recovery backend
- new config block: RADOS_KV
- new option: ceph_conf, userid, pool
Change-Id: Id44afa70e8b5adb2cb2b9d48a807b0046f604f30
Signed-off-by: Gui Hecheng <guihech...@cmss.chinamobile.com
<mailto:guihech...@cmss.chinamobile.com>>
commit f7a09d87851f64a68c2438fdc09372703bcbebec
Author: Matt Benjamin <mbenja...@redhat.com
<mailto:mbenja...@redhat.com>>
AuthorDate: Thu Jul 20 15:21:00 2017 -0400
Commit: Frank S. Filz <ffilz...@mindspring.com
<mailto:ffilz...@mindspring.com>>
CommitDate: Thu Aug 17 14:46:29 2017 -0700
config: add config_url and RADOS url provider
Provides a mechanism to to load nfs-ganesha config sections (e.g.,
export blocks) from a generic URL. Includes a URL provider
which maps URLs to Ceph RADOS objects.
Change-Id: I9067eaef2b38a78e9f1a877dfb9eb3c176239e71
Signed-off-by: Matt Benjamin <mbenja...@redhat.com
<mailto:mbenja...@redhat.com>>
commit b6ce63479c965c12d2d3417abd1dd082cf0967b8
Author: Matt Benjamin <mbenja...@redhat.com
<mailto:mbenja...@redhat.com>>
AuthorDate: Fri Sep 22 14:21:46 2017 -0400
Commit: Frank S. Filz <ffilz...@mindspring.com
<mailto:ffilz...@mindspring.com>>
CommitDate: Fri Sep 22 14:06:12 2017 -0700
rpm spec: add RADOS_URLS
Change-Id: I60ebd4cb5bc3b3184704b8951a5392ed91846cdd
Signed-off-by: Matt Benjamin <mbenja...@redhat.com
<mailto:mbenja...@redhat.com>>
commit 247c4a61cd743e7b3430bb0a9780c3f6d3f73a44
Author: Matt Benjamin <mbenja...@redhat.com
<mailto:mbenja...@redhat.com>>
AuthorDate: Fri Sep 22 15:38:37 2017 -0400
Commit: Frank S. Filz <ffilz...@mindspring.com
<mailto:ffilz...@mindspring.com>>
CommitDate: Fri Sep 22 14:06:28 2017 -0700
rados url: handle error from rados_read()
Change-Id: If437a989ddaea108216c28af99fab6da0f089e01
Signed-off-by: Matt Benjamin <mbenja...@redhat.com
<mailto:mbenja...@redhat.com>>
commit d9f0536b7f3cbe6b9b4d0dc5b4e4acd3337d41b5
Author: Jeff Layton <jlay...@redhat.com <mailto:jlay...@redhat.com>>
AuthorDate: Fri Oct 6 14:23:23 2017 -0400
Commit: Frank S. Filz <ffilz...@mindspring.com
<mailto:ffilz...@mindspring.com>>
CommitDate: Fri Oct 6 14:26:45 2017 -0700
FSAL_CEPH: don't clobber the return code with the getlk call
If a lock is denied, the code will call getlk to get the
conflicting lock
info. That action then clobbers the return code and makes the
lock appear
to be a success.
Also, no need to check conflicting_lock twice here.
See: https://github.com/nfs-ganesha/nfs-ganesha/issues/205
<https://github.com/nfs-ganesha/nfs-ganesha/issues/205>
Change-Id: Ibfc8ca92bec84518573f425131ce969479ae15dd
Signed-off-by: Jeff Layton <jlay...@redhat.com
<mailto:jlay...@redhat.com>>
commit 13a2f2dce7aff5cc86bdb96b058cbb4d20898b66
Author: Matt Benjamin <mbenja...@redhat.com
<mailto:mbenja...@redhat.com>>
AuthorDate: Thu Oct 19 12:58:27 2017 -0400
Commit: Frank S. Filz <ffilz...@mindspring.com
<mailto:ffilz...@mindspring.com>>
CommitDate: Fri Oct 20 15:31:08 2017 -0700
rgw: it's rados_url_param, dude
Amazingly, a one-character mistake broke parsing of the RADOS_URLS
config block.
Change-Id: I6a5661a85cf734e5d93f8c03a78d50ac9a854954
Signed-off-by: Matt Benjamin <mbenja...@redhat.com
<mailto:mbenja...@redhat.com>>
commit 3571d1af1d7b8cb4e9898ed9f92c09ba8fca2805
Author: Matt Benjamin <mbenja...@redhat.com
<mailto:mbenja...@redhat.com>>
AuthorDate: Tue Oct 17 17:24:41 2017 -0400
Commit: Frank S. Filz <ffilz...@mindspring.com
<mailto:ffilz...@mindspring.com>>
CommitDate: Fri Oct 20 15:30:46 2017 -0700
RGW: pass V3 flag to rgw_write() when !state
Needed to permit librgw to simulate opens for NFSv3.
Change-Id: Ia7ba289a5941bd2f68c3ed678a95b9d16cb4fad0
Signed-off-by: Matt Benjamin <mbenja...@redhat.com
<mailto:mbenja...@redhat.com>>
commit 2223c3c70fa2189185e9e968a2ffe0cf763b3544
Author: Matt Benjamin <mbenja...@redhat.com
<mailto:mbenja...@redhat.com>>
AuthorDate: Fri Oct 13 16:55:20 2017 -0400
Commit: Frank S. Filz <ffilz...@mindspring.com
<mailto:ffilz...@mindspring.com>>
CommitDate: Fri Oct 20 14:52:22 2017 -0700
RGW: set fso_compute_readdir_cookie
The change that implement the method, didn't set the fs option.
Change-Id: I9d8e949d08eea8bb86e3a9a40514e52c927bcb26
Signed-off-by: Matt Benjamin <mbenja...@redhat.com
<mailto:mbenja...@redhat.com>>
commit 8d039cd139ea1dcfbe316b7e93e96efe755669ef
Author: Matt Benjamin <mbenja...@redhat.com
<mailto:mbenja...@redhat.com>>
AuthorDate: Tue Oct 17 08:57:57 2017 -0400
Commit: Frank S. Filz <ffilz...@mindspring.com
<mailto:ffilz...@mindspring.com>>
CommitDate: Fri Oct 20 14:52:38 2017 -0700
RGW: use new whence_is_name for READDIR continuation
As of version 1.1.6, librgw provides rgw_readdir2, whose
marker/offset is a dirent name.
Change-Id: I165b34d977b3ed9a2680ee90862b3510d375ee06
Signed-off-by: Matt Benjamin <mbenja...@redhat.com
<mailto:mbenja...@redhat.com>>
commit dfdeb5bae39cf19155ac72e4adff1dd14d8d6759
Author: Frank S. Filz <ffilz...@mindspring.com
<mailto:ffilz...@mindspring.com>>
AuthorDate: Fri Oct 13 09:25:36 2017 -0700
Commit: Frank S. Filz <ffilz...@mindspring.com
<mailto:ffilz...@mindspring.com>>
CommitDate: Wed Oct 18 15:52:48 2017 -0700
MDCACHE: add debug for readdir processing
Change-Id: I9564fbb2c7561c79977da188aaefd013a70f4ce2
Signed-off-by: Frank S. Filz <ffilz...@mindspring.com
<mailto:ffilz...@mindspring.com>>
commit 71833a848499e150aae5453260748d851ad33f93
Author: Frank S. Filz <ffilz...@mindspring.com
<mailto:ffilz...@mindspring.com>>
AuthorDate: Wed Oct 18 13:23:50 2017 -0700
Commit: Frank S. Filz <ffilz...@mindspring.com
<mailto:ffilz...@mindspring.com>>
CommitDate: Wed Oct 18 15:52:49 2017 -0700
Improve readdir debug in upper layers
Change-Id: I3fbaef32b10f2ee6ed3ef7b9bce5ff0a3de175e1
Signed-off-by: Frank S. Filz <ffilz...@mindspring.com
<mailto:ffilz...@mindspring.com>>
commit 95e21745a60d1cd565352af93ddbdf592b501305
Author: Frank S. Filz <ffilz...@mindspring.com
<mailto:ffilz...@mindspring.com>>
AuthorDate: Fri Sep 22 14:01:13 2017 -0700
Commit: Frank S. Filz <ffilz...@mindspring.com
<mailto:ffilz...@mindspring.com>>
CommitDate: Wed Oct 18 15:52:49 2017 -0700
FSAL_VFS: Additional debug messages
Change-Id: I638772aa558afe3acf4a297b3c504ba1da1511db
Signed-off-by: Frank S. Filz <ffilz...@mindspring.com
<mailto:ffilz...@mindspring.com>>
commit 011f90be782b755c763e81783cdd96e7cb48828d
Author: Frank S. Filz <ffilz...@mindspring.com
<mailto:ffilz...@mindspring.com>>
AuthorDate: Tue Oct 17 16:57:12 2017 -0700
Commit: Frank S. Filz <ffilz...@mindspring.com
<mailto:ffilz...@mindspring.com>>
CommitDate: Thu Oct 19 16:26:26 2017 -0700
MDCACHE: Additional cleanup of chunk before reuse or free
Change-Id: I894624fef5a69688dadb0ad0eaa4041f4dd6ab38
Signed-off-by: Frank S. Filz <ffilz...@mindspring.com
<mailto:ffilz...@mindspring.com>>
commit 79dc970258df0bec21ca838f56270628195a2e21
Author: Frank S. Filz <ffilz...@mindspring.com
<mailto:ffilz...@mindspring.com>>
AuthorDate: Tue Oct 17 16:58:48 2017 -0700
Commit: Frank S. Filz <ffilz...@mindspring.com
<mailto:ffilz...@mindspring.com>>
CommitDate: Thu Oct 19 16:26:27 2017 -0700
MDCACHE: Fixup whence_is_name
- Need to distinguish if FSAL readdir is called to fill chunks
to be
used for READDIR response or being called to backfill in
search of the
READDIR cookie/whence. The flag whence_is_search indicates a
cookie is
being searched for, otherwise, chunks are being filled to
fulfill a
READDIR response.
- While continuing backfill with a new chunk, need to update
state.dir_state (the chunk being worked on).
Change-Id: Ie17a1803264cfd41e0d7cb78562e73288dc5ba80
Signed-off-by: Frank S. Filz <ffilz...@mindspring.com
<mailto:ffilz...@mindspring.com>>
commit 56142298c471663c286f4df5527ea4f20feaa26b
Author: Frank S. Filz <ffilz...@mindspring.com
<mailto:ffilz...@mindspring.com>>
AuthorDate: Thu Oct 19 09:46:59 2017 -0700
Commit: Frank S. Filz <ffilz...@mindspring.com
<mailto:ffilz...@mindspring.com>>
CommitDate: Thu Oct 19 16:26:27 2017 -0700
MDCACHE: Don't test MDCACHE_DIR_POPULATED in
mdcache_readdir_chunked
MDCACHE_DIR_POPULATED is ONLY used to determine if we can serve
negative
lookups, which we can only do when we know we have the entire
directory
cached.
Change-Id: I71a0a2959f67f0ac6b64e6b22bc8364901c23433
Signed-off-by: Frank S. Filz <ffilz...@mindspring.com
<mailto:ffilz...@mindspring.com>>
commit c6cdd1b1100dc236d2f668a6c6fac89e7088d8ab
Author: Matt Benjamin <mbenja...@redhat.com
<mailto:mbenja...@redhat.com>>
AuthorDate: Mon Oct 23 15:16:47 2017 -0400
Commit: Frank S. Filz <ffilz...@mindspring.com
<mailto:ffilz...@mindspring.com>>
CommitDate: Fri Oct 27 13:27:31 2017 -0700
RADOS config url: late init RADOS parameters
Initializing RADOS early, before config parsing has started,
prevents connection parameters in the RADOS_URLS config section
from taking effect.
Change-Id: Ied11473806a1b951c05f55b164936f0a1484cd0e
Signed-off-by: Matt Benjamin <mbenja...@redhat.com
<mailto:mbenja...@redhat.com>>
commit 824cea60caab1c37908d5f358da7570de76b43a9
Author: Matt Benjamin <mbenja...@redhat.com
<mailto:mbenja...@redhat.com>>
AuthorDate: Mon Oct 23 16:27:20 2017 -0400
Commit: Frank S. Filz <ffilz...@mindspring.com
<mailto:ffilz...@mindspring.com>>
CommitDate: Fri Oct 27 13:27:07 2017 -0700
RGW: don't deref NULL at LogFullDebug after add_detached_dirent()
The fault happens only at LogFullDebug, as stated.
Change-Id: I64f8688114025237c6d2985ad359e148a0a55737
Signed-off-by: Matt Benjamin <mbenja...@redhat.com
<mailto:mbenja...@redhat.com>>
commit 2ebc97065a5474dfb689b2c7053e955a5e5a3595
Author: Matt Benjamin <mbenja...@redhat.com
<mailto:mbenja...@redhat.com>>
AuthorDate: Tue Oct 24 16:08:40 2017 -0400
Commit: Frank S. Filz <ffilz...@mindspring.com
<mailto:ffilz...@mindspring.com>>
CommitDate: Fri Oct 27 13:27:44 2017 -0700
RGW: REALLY early init support
Provide a general mechanism to access already-identified
config blocks for parsing by later blocks--and use the
mechanism to permit the RADOS config url provider to parse
config variables form its own RADOS_URLS block.
Change-Id: I1e81eae784de09abb3d1fbf715bb05eac6e8333e
Signed-off-by: Matt Benjamin <mbenja...@redhat.com
<mailto:mbenja...@redhat.com>>
commit c4cccca05dedc13de3bae73b17b040be147ca520
Author: Daniel Gryniewicz <d...@redhat.com <mailto:d...@redhat.com>>
AuthorDate: Tue Oct 24 13:07:40 2017 -0400
Commit: Daniel Gryniewicz <d...@redhat.com <mailto:d...@redhat.com>>
CommitDate: Tue Oct 24 13:07:40 2017 -0400
Plumb through Bind_Addr so that it works.
During the conversion to work with IPv6, the code to implement
Bind_Addr
was lost. Implement this code, so that an address can be
given. The
following formats are accepted:
Bind_Addr = aaa.bbb.ccc.ddd; # Standard dotted quad IPv4 address
Bind_Addr = a:b:c:d:e:f:g:h; # Standard IPv6 address (may have ::)
Bind_Addr = ::FFFF:aaa.bbb.ccc.ddd; # IPv4 address mapped in IPv6
Bind_Addr = 0.0.0.0; # Listen on all addresses (IPv4 and IPv6)
Bind_Addr = ::; # Listen on all IPv6 addresses (no IPv4)
Note that none of these are quoted in any way. Case of hex
does not
matter.
Change-Id: Ie960e8325bdabb3c307dff03694f3f1dc92f4596
Signed-off-by: Daniel Gryniewicz <d...@redhat.com
<mailto:d...@redhat.com>>
------------------------------------------------------------------------------
Check out the vibrant tech community on one of the world's most
engaging tech sites, Slashdot.org! http://sdm.link/slashdot
_______________________________________________
Nfs-ganesha-devel mailing list
Nfs-ganesha-devel@lists.sourceforge.net
<mailto:Nfs-ganesha-devel@lists.sourceforge.net>
https://lists.sourceforge.net/lists/listinfo/nfs-ganesha-devel
<https://lists.sourceforge.net/lists/listinfo/nfs-ganesha-devel>
------------------------------------------------------------------------------
Check out the vibrant tech community on one of the world's most
engaging tech sites, Slashdot.org! http://sdm.link/slashdot
_______________________________________________
Nfs-ganesha-devel mailing list
Nfs-ganesha-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/nfs-ganesha-devel
