glusterfs.git/xlators/cluster/ec/src/ec.h, branch v5.2

Land clang-format changes

2018-09-12T11:52:48+00:00

Change-Id: I6f5d8140a06f3c1b2d196849299f8d483028d33b

cluster/ec: Fix possible shift overflow

2017-12-22T16:19:53+00:00

A coverity scan has revelaed a potential shift overflow while scanning
the bitmap of available subvolumes. The actual overflow cannot happen,
but I've changed to test used to control the limit to make it explicit.

Change-Id: Ieb55f010bbca68a1d86a93e47822f7c709a26e83
BUG: 789278
Signed-off-by: Xavier Hernandez

cluster/ec: Keep last written strip in in-memory cache

2017-11-10T22:15:37+00:00

Problem:
Consider an EC volume with configuration  4 + 2.
The stripe size for this would be 512 * 4 = 2048.
That means, 2048 bytes of user data stored in one
stripe. Let's say 2048 + 512 = 2560 bytes are
already written on this volume. 512 Bytes would
be in second stripe. Now, if there are sequential
writes with offset 2560 and of size 1 Byte, we have
to read the whole stripe, encode it with 1 Byte and
then again have to write it back. Next, write with
offset 2561 and size of 1 Byte will again
READ-MODIFY-WRITE the whole stripe. This is causing
bad performance because of lots of READ request
travelling over the network.

There are some tools and scenario's where such kind
of load is coming and users are not aware of that.
Example: fio and zip

Solution:
One possible solution to deal with this issue is to
keep last stripe in memory. This way, we need not to
read it again and we can save READ fop going over the
network. Considering the above example, we have to
keep last 2048 bytes (maximum) in memory per file.

Change-Id: I3f95e6fc3ff81953646d374c445a40c6886b0b85
BUG: 1471753
Signed-off-by: Ashish Pandey

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: Provide an option to enable/disable eager lock

2016-03-16T04:54:28+00:00

Problem: If a fop takes lock, and completes its operation,
it waits for 1 second before releasing the lock. However,
If ec find any lock contention within this time period,
it release the lock immediately before time expires. As we
take lock on first brick, for few operations, like read, it
might happen that discovery of lock contention might take
long time and can degrades the performance.

Solution: Provide an option to enable/disable eager lock.
If eager lock is disabled, lock will be released as soon
as fop completes.

gluster v set  disperse.eager-lock on
gluster v set  disperse.eager-lock off

Change-Id: I000985a787eba3c190fdcd5981dfbf04e64af166
BUG: 1314649
Signed-off-by: Ashish Pandey 
Reviewed-on: http://review.gluster.org/13605
Smoke: Gluster Build System 
Reviewed-by: Pranith Kumar Karampuri 
Tested-by: Pranith Kumar Karampuri 
CentOS-regression: Gluster Build System 
NetBSD-regression: NetBSD Build System

cluster/ec: Create this->itable in all cases

2016-01-29T05:12:53+00:00

Problem:
glfsheal operates based on mount's volfile which doesn't have iamshd flag due
to which this->itable is NULL, this leads to "inode not found" logs in glfsheal
logs.

Fix:
Ec only allocates itable with 10 inodes, so allocating this->itable in all
cases in init.

Change-Id: I01d3c05e93a17007a4716a2d6f392d2aa306a34b
BUG: 1294743
Signed-off-by: Pranith Kumar K 
Reviewed-on: http://review.gluster.org/13112
Smoke: Gluster Build System 
NetBSD-regression: NetBSD Build System 
CentOS-regression: Gluster Build System

cluster/ec: Implement gfid-hash read-policy

2015-10-09T12:26:05+00:00

Add a policy in ec to performs reads from same bricks as long as they
are good. Based on the gfid of the file/directory it determines the
bricks to be considered for reading.

Change-Id: Ic97b5c54c086a28b5e07a330a4fd448551b49376
BUG: 1261260
Signed-off-by: Pranith Kumar K 
Reviewed-on: http://review.gluster.org/12133
Tested-by: NetBSD Build System 
Tested-by: Gluster Build System 
Reviewed-by: Xavier Hernandez

cluster/ec: Make background healing optional behavior

2015-07-06T10:31:27+00:00

Provide options to control number of active background heal count and qlen.

Change-Id: Idc2419219d881f47e7d2e9bbc1dcdd999b372033
BUG: 1237381
Signed-off-by: Pranith Kumar K 
Reviewed-on: http://review.gluster.org/11473
Reviewed-by: Xavier Hernandez 
Tested-by: Gluster Build System

cluster/ec: Add throttling in background healing

2015-07-02T00:10:00+00:00

- 8 parallel heals can happen.
- 128 heals will wait for their turn
- Heals will be rejected if 128 heals are already waiting.

Change-Id: I2e99bf064db7bce71838ed9901a59ffd565ac390
BUG: 1237381
Signed-off-by: Pranith Kumar K 
Reviewed-on: http://review.gluster.org/11471
Tested-by: Gluster Build System 
Reviewed-by: Xavier Hernandez 
Tested-by: NetBSD Build System

cluster/ec: Correctly cleanup delayed locks

2015-05-20T08:00:01+00:00

When a delayed lock is pending, a graph switch doesn't correctly
terminate it. This means that the update of version and size xattrs
is lost, causing EIO errors.

This patch handles GF_EVENT_PARENT_DOWN event to correctly finish
pending udpdates before completing the graph switch.

Change-Id: I394f3b8d41df8d83cdd36636aeb62330f30a66d5
BUG: 1188145
Signed-off-by: Xavier Hernandez 
Reviewed-on: http://review.gluster.org/10787
Tested-by: NetBSD Build System
Tested-by: Gluster Build System 
Reviewed-by: Pranith Kumar Karampuri