glusterfs.git/xlators/nfs/server, branch v3.12dev

nfs: Check for null buf, and set op_errno to EIO not 0

2017-04-26T23:31:45+00:00

Summary:
- Prevent crashes for the case where "getattr" actually failed and
  returned with a NULL buf, but the op_errno was set to 0.
- This is a cherry-pick of D1571184

Signed-off-by: Shreyas Siravara 
Change-Id: Ia2d6ad7539df714f9420dcf063c7c14e727bb7e3
BUG: 1428064
Reviewed-on: http://review.gluster.org/16152
Smoke: Gluster Build System 
NetBSD-regression: NetBSD Build System 
CentOS-regression: Gluster Build System 
Reviewed-by: Kevin Vigor 
Reviewed-on: https://review.gluster.org/16803
Reviewed-by: Amar Tumballi 
Tested-by: Jeff Darcy 
Reviewed-by: Jeff Darcy

nfs: Tear down transports for requests that arrive before the volume is initialized

2017-04-26T23:31:21+00:00

Summary:
- Disconnects RPC transports for requests that cannot be serviced
  because volumes are not ready.
- This is a cherry-pick of D2991403

Signed-off-by: Shreyas Siravara 
Change-Id: I07ff0795b81d25624541ff981b5f2586d078e9a6
BUG: 1428068
Reviewed-on: http://review.gluster.org/16154
Smoke: Gluster Build System 
NetBSD-regression: NetBSD Build System 
Reviewed-by: Kevin Vigor 
CentOS-regression: Gluster Build System 
Reviewed-on: https://review.gluster.org/16805
Reviewed-by: Amar Tumballi 
Tested-by: Jeff Darcy 
Reviewed-by: Jeff Darcy

nfs: Fix compiler warning when calling svc_getcaller

2017-04-11T15:25:37+00:00

Summary:
When using libtirpc instead of glibc rpc the result needs to be
cast to (struct sockaddr_in *)

This diff was originally a cherry-pick of D3111554 to 3.8

Signed-off-by: Kaleb S. KEITHLEY 
Signed-off-by: Shreyas Siravara 
Change-Id: If4c27dbe6c032f9e278ea08cd3c96a4d07bcc5f9
BUG: 1428073
Reviewed-on: http://review.gluster.org/16179
Tested-by: Shreyas Siravara 
NetBSD-regression: NetBSD Build System 
Smoke: Gluster Build System 
Reviewed-by: Kevin Vigor 
CentOS-regression: Gluster Build System 
Reviewed-on: https://review.gluster.org/16804
Reviewed-by: Amar Tumballi 
Reviewed-by: Niels de Vos

nfs: make subdir mounting work for Solaris 10 clients

2017-04-07T10:37:28+00:00

This fixes the segfault caused by solaris client in Gluster/NFS.
Volname was not being parsed properly, Instead of volume
name complete path was being used in nfs_mntpath_to_xlator().
Fixed it by striping volume name from complete path  in nfs_mntpath_to_xlator().

Modified function name nfs3_funge_solaris_zerolen_fh() to
nfs3_funge_webnfs_zerolen_fh() as zero-filled filehandle is specific to WebNFS.
RFC : https://tools.ietf.org/html/rfc2055
Solaris uses WebNFS, the zero-filled FH is defined in the WebNFS spec.

Logic was even added in fuction nfs3_funge_webnfs_zerolen_fh() to send
subdir path in function glfs_resolve_at() instead of complete path for
subdir mount.

Change-Id: I19aae3547b8910e7ed4974ee5385424cab3e834a
BUG: 1426667
Signed-off-by: Bipin Kunal 
Reviewed-on: https://review.gluster.org/16770
Reviewed-by: Niels de Vos 
Smoke: Gluster Build System 
NetBSD-regression: NetBSD Build System 
CentOS-regression: Gluster Build System

nfs: use pkill on all OS's

2017-03-27T11:57:04+00:00

'killall' is the only executable that is needed from the psmisc package
on glusterfs-server installations. NetBSD already uses 'pkill' as
alternative, this can be used everywhere.

BUG: 1197308
Change-Id: Ied671a74b735e321f52d99c869627b957aa2bd02
Signed-off-by: Niels de Vos 
Reviewed-on: https://review.gluster.org/16928
Smoke: Gluster Build System 
NetBSD-regression: NetBSD Build System 
CentOS-regression: Gluster Build System 
Reviewed-by: Kaleb KEITHLEY

nfs: remove include of mount3.h

2017-02-23T13:06:28+00:00

While there is proper protection against that, coverity
complain of a loop, and nothing in auth-cache.h use
a declaration of "mount3.h".

Change-Id: I55885ec12496359170f271ae192e9940a35121e1
BUG: 789278
Signed-off-by: Michael Scherer 
Reviewed-on: https://review.gluster.org/16728
Reviewed-by: Niels de Vos 
Tested-by: Niels de Vos 
Smoke: Gluster Build System 
NetBSD-regression: NetBSD Build System 
CentOS-regression: Gluster Build System

gNFS: Keep the mountdict as long as the service is active

2017-02-14T09:16:46+00:00

We initialize and take ref once on mountdict during NFS/MNT3 server
initialization but seem to be unref'in it for every UMNTALL request.
This can lead to crash when there are multiple UMNTALL requests
with >=1 active mount entry(/ies) in the mountlist.

Since we take the ref only once, we should keep the mountdict through
out the life of the process and dereference it only during unitialization
of mnt3 service.

Change-Id: I3238a8df09b8972e56dd93fee426d866d40d9959
BUG: 1421759
Signed-off-by: Soumya Koduri 
Reviewed-on: https://review.gluster.org/16611
Smoke: Gluster Build System 
Reviewed-by: Raghavendra Bhat 
NetBSD-regression: NetBSD Build System 
CentOS-regression: Gluster Build System 
Reviewed-by: jiffin tony Thottan 
Reviewed-by: Niels de Vos

core: run many bricks within one glusterfsd process

2017-01-31T00:13:58+00:00

This patch adds support for multiple brick translator stacks running
in a single brick server process.  This reduces our per-brick memory usage by
approximately 3x, and our appetite for TCP ports even more.  It also creates
potential to avoid process/thread thrashing, and to improve QoS by scheduling
more carefully across the bricks, but realizing that potential will require
further work.

Multiplexing is controlled by the "cluster.brick-multiplex" global option.  By
default it's off, and bricks are started in separate processes as before.  If
multiplexing is enabled, then *compatible* bricks (mostly those with the same
transport options) will be started in the same process.

Change-Id: I45059454e51d6f4cbb29a4953359c09a408695cb
BUG: 1385758
Signed-off-by: Jeff Darcy 
Reviewed-on: https://review.gluster.org/14763
Smoke: Gluster Build System 
NetBSD-regression: NetBSD Build System 
CentOS-regression: Gluster Build System 
Reviewed-by: Vijay Bellur

gNFS : pass proper gfid for nfs_setattr() in nfs3svc_create_cbk()

2017-01-30T11:51:51+00:00

The nfs3svc_create_cbk() takes "gfid" from inode variable which it
received, not from inode which it performed linking to nfs_setattr().
There is possiblity that the inode passed into this function holds
NULL gfid if there are some parallel operation happening on the same
file.

Similar issue is mentioned in https://bugzilla.redhat.com/show_bug.cgi?id=1413971
Thanks pranith for proposing the fix

Change-Id: I1a0ff4f02b483416f19f4f064c306c2cad5d9d8b
BUG: 1413971
Signed-off-by: Jiffin Tony Thottan 
Reviewed-on: https://review.gluster.org/16421
Reviewed-by: soumya k 
NetBSD-regression: NetBSD Build System 
CentOS-regression: Gluster Build System 
Smoke: Gluster Build System 
Reviewed-by: Niels de Vos

refcount: return pointer to the structure instead of a counter

2016-12-11T08:47:05+00:00

There are no real users of the counter. It was thought of a handy tool
to track and debug refcounting, but it is not used at all. Some parts of
the code would benefit from a pointer getting returned instead.

BUG: 1399780
Change-Id: I97e52c48420fed61be942ea27ff4849b803eed12
Signed-off-by: Niels de Vos 
Reviewed-on: http://review.gluster.org/15971
Smoke: Gluster Build System 
NetBSD-regression: NetBSD Build System 
CentOS-regression: Gluster Build System 
Reviewed-by: Xavier Hernandez 
Reviewed-by: Kaleb KEITHLEY