glusterfs.git/xlators/features/shard/src/shard.h, branch v3.11.0beta1

features/shard: Fix vm corruption upon fix-layout

2017-04-07T18:52:00+00:00

shard's writev implementation, as part of identifying
presence of participant shards that aren't in memory,
first sends an MKNOD on these shards, and upon EEXIST error,
looks up the shards before proceeding with the writes.

The VM corruption was caused when the following happened:
1. DHT had n subvolumes initially.
2. Upon add-brick + fix-layout, the layout of .shard changed
   although the existing shards under it were yet to be migrated
   to their new hashed subvolumes.
3. During this time, there were writes on the VM falling in regions
   of the file whose corresponding shards were already existing under
   .shard.
4. Sharding xl sent MKNOD on these shards, now creating them in their
   new hashed subvolumes although there already exist shard blocks for
   this region with valid data.
5. All subsequent writes were wound on these newly created copies.

The net outcome is that both copies of the shard didn't have the correct
data. This caused the affected VMs to be unbootable.

FIX:
For want of better alternatives in DHT, the fix changes shard fops to do
a LOOKUP before the MKNOD and upon EEXIST error, perform another lookup.

Change-Id: I8a2e97d91ba3275fbc7174a008c7234fa5295d36
BUG: 1440051
RCA'd-by: Raghavendra Gowdappa 
Reported-by: Mahdi Adnan 
Signed-off-by: Krutika Dhananjay 
Reviewed-on: https://review.gluster.org/17010
Smoke: Gluster Build System 
Reviewed-by: Pranith Kumar Karampuri 
NetBSD-regression: NetBSD Build System 
CentOS-regression: Gluster Build System

features/shard: Fix EIO error on add-brick

2017-02-24T04:57:43+00:00

DHT seems to link inode during lookup even before initializing
inode ctx with layout information, which comes after
directory healing.

Consider two parallel writes. As part of the first write,
shard sends lookup on .shard which in its return path would
cause DHT to link .shard inode. Now at this point, when a
second write is wound, inode_find() of .shard succeeds and
as a result of this, shard goes to create the participant
shards by issuing MKNODs under .shard. Since the layout is
yet to be initialized, mknod fails in dht call path with EIO,
leading to VM pauses.

The fix involves shard maintaining a flag to denote whether
a fresh lookup on .shard completed one network trip. If it
didn't, all inode_find()s in fop path will be followed by a
lookup before proceeding with the next stage of the fop.

Big thanks to Raghavendra G and Pranith Kumar K for the RCA
and subsequent inputs and feedback on the patch.

Change-Id: I9383ec7e3f24b34cd097a1b01ad34e4eeecc621f
BUG: 1420623
Signed-off-by: Krutika Dhananjay 
Reviewed-on: https://review.gluster.org/14419
Reviewed-by: Pranith Kumar Karampuri 
NetBSD-regression: NetBSD Build System 
CentOS-regression: Gluster Build System 
Smoke: Gluster Build System 
Reviewed-by: Atin Mukherjee

features/shard: Put onus of choosing the inode to resolve on individual fops

2017-02-23T07:56:38+00:00

... as opposed to adding checks in "common" functions to choose the inode
to resolve based local->fop, which is rather ugly and prone to errors.

Change-Id: Ia46cc59992baa2979516369cb72d8991452c0274
BUG: 1420623
Signed-off-by: Krutika Dhananjay 
Reviewed-on: https://review.gluster.org/16709
Tested-by: Pranith Kumar Karampuri 
NetBSD-regression: NetBSD Build System 
CentOS-regression: Gluster Build System 
Smoke: Gluster Build System 
Reviewed-by: Pranith Kumar Karampuri

features/changelog: Capture FXATTROP and XATTROP in changelog

2015-11-23T06:49:15+00:00

GEO-REP INTEROP WITH SHARD FEATURE

shard xlator updates size of the file using FXATTROP
or XATTROP. Hence record the same in changelog.

Change-Id: Ie0c21e9326da05ea78dc1ef3fd32a90ef38b4bb9
BUG: 1265148
Signed-off-by: Kotresh HR 
Reviewed-on: http://review.gluster.org/12225
Tested-by: NetBSD Build System 
Tested-by: Gluster Build System 
Reviewed-by: Venky Shankar

features/shard: Force cache-refresh when lookup/readdirp/stat detect that xattr value has changed

2015-10-28T13:38:35+00:00

Change-Id: Ia3225a523287f6689b966ba4f893fc1b1fa54817
BUG: 1272986
Signed-off-by: Krutika Dhananjay 
Reviewed-on: http://review.gluster.org/12400
Reviewed-by: Pranith Kumar Karampuri 
Tested-by: Gluster Build System

features/shard: Regulate memory consumption by individual shards' inode_t objects

2015-10-09T05:17:13+00:00

Shard translator will now maintain an lru list of inodes associated with
individual shards of constant size, and will make sure that at no point the
number of these inodes will exceed the configured limit.
This is to keep the memory consumption by the thousands of shards of every large
file from exploding.

Change-Id: I5e60eea5dcf3130257fb431ca70cfaba53cae7f3
BUG: 1252263
Signed-off-by: Krutika Dhananjay 
Reviewed-on: http://review.gluster.org/12254
Tested-by: NetBSD Build System 
Reviewed-by: Pranith Kumar Karampuri

features/shard: Port log messages to new framework

2015-09-28T06:08:36+00:00

Change-Id: Iac01e6a89a0d0c37a12a5e47f17f7ced85a31590
BUG: 1265516
Signed-off-by: Krutika Dhananjay 
Reviewed-on: http://review.gluster.org/12217
Tested-by: NetBSD Build System 
Tested-by: Gluster Build System 
Reviewed-by: Pranith Kumar Karampuri

features/shard: Performance improvements in IO path

2015-09-23T05:51:38+00:00

This is patch 1/2 of the performance improvement work
for sharding in the IO path.
What this patch does:
Since the primary use-case where sharding is targeted -
VM store - is a single-writer workload, instead of
performing lookup on the base file everytime to gather the
size and block count from the backend in reads, writes and
truncate, now the size and block count is also cached and
kept up-to-date after every inode write in the inode ctx.

TO-DO:
Make changes in rename, link, unlink, [f]setattr and [f]stat
to keep the relevant iatt members up-to-date in the inode ctx.

Change-Id: Ica87d020dabc3a3dbccec814b26b01d6a629ff4d
BUG: 1258905
Signed-off-by: Krutika Dhananjay 
Reviewed-on: http://review.gluster.org/12126
Tested-by: NetBSD Build System 
Tested-by: Gluster Build System 
Reviewed-by: Pranith Kumar Karampuri

features/shard: Filter internal shard xattrs in {get,remove,set}xattr

2015-09-09T02:14:10+00:00

Change-Id: I40e4a5dbd13d6c3d777e7e01f93dabc83e52b137
BUG: 1260637
Signed-off-by: Krutika Dhananjay 
Reviewed-on: http://review.gluster.org/12121
Reviewed-by: Pranith Kumar Karampuri 
Tested-by: NetBSD Build System

features/shard: Fix permission issues

2015-08-30T15:20:21+00:00

This patch does the following:
* reverts commit b467af0e99b39ef708420d3f7f6696b0ca618512
* changes ownership on shards under /.shard to be root:root
* makes readv, writev, [f]truncate, rename, and unlink fops
  to perform operations on files under /.shard with
  frame->root->{uid,gid} as 0.

This would ensure that a [f]setattr on a sharded file
does not need to be called on all the shards associated with it.

Change-Id: Idcfb8c0dd354b0baab6b2356d2ab83ce51caa20e
BUG: 1251824
Signed-off-by: Krutika Dhananjay 
Reviewed-on: http://review.gluster.org/11992
Reviewed-by: Pranith Kumar Karampuri 
Tested-by: NetBSD Build System 
Tested-by: Gluster Build System