glusterfs.git/xlators/cluster/ec/src/ec-inode-write.c, branch v4.0dev

cluster/ec: Implement FALLOCATE FOP for EC

2017-05-23T07:13:06+00:00

FALLOCATE file operations is not implemented in the
existing EC code. This change set implements it
for EC.

BUG: 1448293
Change-Id: Id9ed914db984c327c16878a5b2304a0ea461b623
Signed-off-by: Sunil Kumar Acharya 
Reviewed-on: https://review.gluster.org/15200
Smoke: Gluster Build System 
NetBSD-regression: NetBSD Build System 
Reviewed-by: Pranith Kumar Karampuri 
CentOS-regression: Gluster Build System

cluster/ec: Check xdata to avoid memory leak

2016-12-02T16:24:32+00:00

Problem: ec_writev_start calls ec_make_internal_fop_xdata
to set "yes" in xdata before ec_readv (an internal fop)
is called for head and tail. Second call to this function
is overwriting the previous allocated dict_t to "xdata",
which results in memory leak.

Solution: In ec_make_internal_fop_xdata, check if *xdata
is NULL or not to avoid overwriting *xdata.

Change-Id: I49b83923e11aff9b92d002e86424c0c2e1f5f74f
BUG: 1400818
Signed-off-by: Ashish Pandey 
Reviewed-on: http://review.gluster.org/16007
Reviewed-by: Xavier Hernandez 
Reviewed-by: Pranith Kumar Karampuri 
Tested-by: Pranith Kumar Karampuri 
Smoke: Gluster Build System 
NetBSD-regression: NetBSD Build System 
CentOS-regression: Gluster Build System

cluster/ec: Add support for hardware acceleration

2016-09-08T17:08:25+00:00

This patch implements functionalities for fast encoding/decoding
using hardware support. Currently optimized x86_64, SSE and AVX is
added.

Additionally this patch implements a caching mecanism for inverse
matrices to reduce computation time, as well as a new method for
computing the inverse that takes quadratic time instead of cubic.

Finally some unnecessary memory copies have been eliminated to
further increase performance.

Change-Id: I26c75f26fb4201bd22b51335448ea4357235065a
BUG: 1289922
Signed-off-by: Xavier Hernandez 
Reviewed-on: http://review.gluster.org/12837
Tested-by: Pranith Kumar Karampuri 
Smoke: Gluster Build System 
NetBSD-regression: NetBSD Build System 
CentOS-regression: Gluster Build System 
Reviewed-by: Pranith Kumar Karampuri

cluster/ec: Create copy of dict for setting internal xattrs

2015-12-01T17:21:17+00:00

Problem:
Ec takes a ref of the request xdata and sets trusted.ec.version/algo etc xattrs
as part of it. But this request xdata could be using same dictionary to do the
operation on multiple subvolumes, due to which other subvolumes will have
internal xattrs of ec in it and will be created on subvols where they are not
supposed to appear.

Fix:
Take a copy of the request xdata/dict to prevent this from happening.

Most of the debugging work and test script is contributed by Nitya.

BUG: 1286910
Change-Id: If146435dfb89656158dbed3862a3e9a0cda60581
Signed-off-by: Pranith Kumar K 
Reviewed-on: http://review.gluster.org/12831
Tested-by: NetBSD Build System 
Tested-by: Gluster Build System 
Reviewed-by: Xavier Hernandez

cluster/ec: Mark internal fops appropriately

2015-11-19T09:25:50+00:00

1) Mark read fops in read-modify-write by EC as internal.
2) Handle uid/gid set/reset correctly

BUG: 1282761
Change-Id: I5c1ce0cd6213367eaead5fed33aa2397c4e46df7
Signed-off-by: Xavier Hernandez 
Reviewed-on: http://review.gluster.org/12599
Tested-by: Gluster Build System 
Tested-by: NetBSD Build System 
Reviewed-by: Pranith Kumar Karampuri

fd: Do fd_bind on successful open

2015-08-28T17:26:17+00:00

- fd_unref should decrement fd->inode->fd_count only if it is present in the
inode's fd list.
- successful open/opendir should perform fd_bind.

Change-Id: I81dd04f330e2fee86369a6dc7147af44f3d49169
BUG: 1207735
Signed-off-by: Pranith Kumar K 
Reviewed-on: http://review.gluster.org/11044
Reviewed-by: Anoop C S 
Tested-by: NetBSD Build System 
Tested-by: Gluster Build System 
Reviewed-by: Raghavendra G

cluster/ec: Fix tracking of good bricks

2015-08-06T17:12:22+00:00

The bitmask of good and bad bricks was kept in the context of the
corresponding inode or fd. This was problematic when an external
process (another client or the self-heal process) did heal the
bricks but no one changed the bitmaks of other clients.

This patch removes the bitmask stored in the context and calculates
which bricks are healthy after locking them and doing the initial
xattrop. After that, it's updated using the result of each fop.

Change-Id: I225e31cd219a12af4ca58871d8a4bb6f742b223c
BUG: 1236065
Signed-off-by: Xavier Hernandez 
Reviewed-on: http://review.gluster.org/11844
Tested-by: NetBSD Build System 
Tested-by: Gluster Build System 
Reviewed-by: Pranith Kumar Karampuri

cluster/ec: Minimize usage of EIO error

2015-07-28T11:12:17+00:00

Change-Id: I82e245615419c2006a2d1b5e94ff0908d2f5e891
BUG: 1245276
Signed-off-by: Xavier Hernandez 
Reviewed-on: http://review.gluster.org/11741
Tested-by: Gluster Build System 
Reviewed-by: Pranith Kumar Karampuri 
Tested-by: NetBSD Build System

ec: Porting messages to new logging framework

2015-06-26T15:51:59+00:00

Change-Id: Ia05ae750a245a37d48978e5f37b52f4fb0507a8c
BUG: 1194640
Signed-off-by: Nandaja Varma 
Reviewed-on: http://review.gluster.org/10465
Tested-by: NetBSD Build System 
Reviewed-by: Xavier Hernandez

cluster/ec: Forced unlock when lock contention is detected

2015-05-27T10:25:47+00:00

EC uses an eager lock mechanism to optimize multiple read/write
requests on the same entry or inode. This increases performance
but can have adverse results when other clients try to access the
same entry/inode.

To solve this, this patch adds a functionality to detect when this
happens and force an earlier release to not block other clients.

The method consists on requesting GF_GLUSTERFS_INODELK_COUNT and
GF_GLUSTERFS_ENTRYLK_COUNT for all fops that take a lock. When this
count is greater than one, the lock is marked to be released. All
fops already waiting for this lock will be executed normally before
releasing the lock, but new requests that also require it will be
blocked and restarted after the lock has been released and reacquired
again.

Another problem was that some operations did correctly lock the
parent of an entry when needed, but got the size and version xattrs
from the entry instead of the parent.

This patch solves this problem by binding all queries of size and
version to each lock and replacing all entrylk calls by inodelk ones
to remove concurrent updates on directory metadata.  This also allows
rename to correctly update source and destination directories.

Change-Id: I2df0b22bc6f407d49f3cbf0733b0720015bacfbd
BUG: 1165041
Signed-off-by: Xavier Hernandez 
Reviewed-on: http://review.gluster.org/10852
Tested-by: NetBSD Build System
Tested-by: Gluster Build System 
Reviewed-by: Pranith Kumar Karampuri