glusterfs.git/tests/features, branch v3.11dev

jbr: Sending rollback from failed fop to fdl

2016-11-08T19:25:25+00:00

In case of a failed fop, the failure is detected
by the leader in the jbr-server in two places. First
during a quorum check of +ve responses when it
receives responses from all the followers. At this
point if the fop hasn't been successfully journaled
at a quorum of followers (as in there is no merit in
trying the fop in the leader as the quorum will never
be met), then we fail the fop.

Also if this quorum is met, then the fop is tried on
the leader, and after the leader completes the fop
a quorum check similar to the previous one is done
again, this time including the leaders outcome. If
quorum is not met, then we fail the fop.

In both these cases, when the fop fails we send a -ve
ack to the client. With this patch, now we will also
send a rollback through a GF_FOP_IPC to all the followers(and
also to the leader in the second case of failure). This
rollback will contain the index and term number of the
fop which failed. This will be recorded in the respective
journals of the bricks and will be used to rollback the
fop on that brick later.

A subsequent write, and it's respective rollback would
look something like the following in the journal.

The trusted.jbr.term and trusted.jbr.index present in the
dict of both the logs, relate them, and the presence of
"rollback-fop" in the dict of IPC indicates that it is a
rollback fop, and the value 13(stands for GF_FOP_WRITE)
indicates what kind of rollback operation it is.

=== GF_FOP_WRITE
fd = 
vector = <158 bytes>
offset = 0 (0x0)
flags = 32769 (0x8001)
xdata = dict {
 trusted.jbr.term = 0 <2 bytes>
 trusted.jbr.index = 4 <2 bytes>
}
=== GF_FOP_IPC
xdata = dict {
 trusted.jbr.term = 0 <2 bytes>
 trusted.jbr.index = 4 <2 bytes>
 rollback-fop = 13 <3 bytes>
}

Change-Id: I70b6a143d20697153d58e2f719e34ecd1ed160a5
BUG: 1349385
Signed-off-by: Avra Sengupta 
Reviewed-on: http://review.gluster.org/14783
NetBSD-regression: NetBSD Build System 
CentOS-regression: Gluster Build System 
Reviewed-by: Jeff Darcy 
Smoke: Gluster Build System

tests: disable lock_revocation.t on NetBSD

2016-08-31T12:15:59+00:00

This has been consistently causing hangs in NetBSD machines. I have not
been able to debug the issue and we have merge deadline for 3.9. It
would be better to disable this for now.

Change-Id: I8c63940aa26f78dd9994bb63293a5757835ec52b
BUG: 1369401
Signed-off-by: Raghavendra Talur 
Reviewed-on: http://review.gluster.org/15374
Smoke: Gluster Build System 
CentOS-regression: Gluster Build System 
NetBSD-regression: Gluster Build System 
Reviewed-by: Jeff Darcy

glusterd: clean up old port and allocate new one on every restart

2016-08-04T04:43:34+00:00

GlusterD as of now was blindly assuming that the brick port which was already
allocated would be available to be reused and that assumption is absolutely
wrong.

Solution : On first attempt, we thought GlusterD should check if the already
allocated brick ports are free, if not allocate new port and pass it to the
daemon. But with that approach there is a possibility that if PMAP_SIGNOUT is
missed out, the stale port will be given back to the clients where connection
will keep on failing. Now given the port allocation always start from base_port,
if everytime a new port has to be allocated for the daemons, the port range will
still be under control. So this fix tries to clean up old port using
pmap_registry_remove () if any and then goes for pmap_registry_alloc ()

Change-Id: If54a055d01ab0cbc06589dc1191d8fc52eb2c84f
BUG: 1221623
Signed-off-by: Atin Mukherjee 
Reviewed-on: http://review.gluster.org/15005
Smoke: Gluster Build System 
NetBSD-regression: NetBSD Build System 
CentOS-regression: Gluster Build System 
Reviewed-by: Avra Sengupta

features/locks: Add lock revocation functionality to posix locks translator

2016-07-18T09:30:38+00:00

Summary:
- Motivation: Prevents cluster instability by mis-behaving clients
causing bricks to OOM due to inode/entry lock pile-ups.
- Adds option to strip clients of entry/inode locks after N seconds
- Adds option to clear ALL locks should the revocation threshold get hit
- Adds option to clear all or granted locks should the max-blocked
threshold get hit (can be used in combination w/ revocation-clear-all).
- Options are:
features.locks-revocation-secs
features.locks-revocation-clear-all [on/off]
features.locks-revocation-max-blocked
- Adds monkey-locking option to ignore 1% of unlock requests (dev only)
features.locks-monkey-unlocking [on/off]
- Adds logging to indicate revocation event & reason

Test Plan:
First you will need TWO fuse mounts for this repro. Call them /mnt/patchy1 & /mnt/patchy2.

1. Enable monkey unlocking on the volume:
gluster vol set patchy features.locks-monkey-unlocking on

2. From the "patchy1", use DD or some other utility to begin writing to a file,
eventually the dd will hang due to the dropped unlocked requests. This now
simulates the broken client. Run:

for i in {1..1000};do dd if=/dev/zero of=/mnt/patchy1/testfile bs=1k count=10;done'

...this will eventually hang as the unlock request has been lost.

3. Goto another window and setup the mount "patchy2" @ /mnt/patchy2, and
observe that 'echo "hello" >> /mnt/patchy2/testfile" will hang due to the
inability of the client to take out the required lock.

4. Next, re-start the test this time enabling lock revocation; use a timeout of
2-5 seconds for testing:
'gluster vol set patchy features.locks-revocation-secs <2-5>'

5. Wait 2-5 seconds before executing step 3 above this time. Observe that this
time the access to the file will succeed, and the writes on patchy1 will
unblock until they hit another failed unlock request due to
"monkey-unlocking".

BUG: 1350867
Change-Id: I814b9f635fec53834a26db634d1300d9a61057d8
Signed-off-by: Pranith Kumar K
Reviewed-on: http://review.gluster.org/14816
NetBSD-regression: NetBSD Build System
Reviewed-by: Krutika Dhananjay
CentOS-regression: Gluster Build System
Smoke: Gluster Build System

api: use versioned symbols for minor ABI change

2016-06-15T01:30:21+00:00

No need to expand the API/ABI. E.g., see how glfs_lookupat
changed between 3.7.0 and 3.7.4 IIRC

(I originally argued against versioning the library. I wanted
to just add new functions as they were needed, as was initially
done for glfs_ipc and glfs_ipc_xd in the master branch for 4.0.
But others strongly wanted versioning.)

Having made the decision to use versioning, I believe we should
continue. At least until we have a public decision that we're
no longer going to use versioning.

Change-Id: I0c3b2c1cbb297ae2b2864b647c224922987d74ad
Signed-off-by: Kaleb S KEITHLEY 
Reviewed-on: http://review.gluster.org/14717
Reviewed-by: Shyamsundar Ranganathan 
Smoke: Gluster Build System 
NetBSD-regression: NetBSD Build System 
CentOS-regression: Gluster Build System 
Reviewed-by: Jeff Darcy

glusterd: volgen and volume set changes for leases

2016-06-03T04:58:43+00:00

Add leases xlator in volgen and also add corresponding volume set options

Change-Id: Ic5de50cdb87eaf6a833e739bc7e08fecbeca3de3
BUG: 1319992
Signed-off-by: Poornima G 
Reviewed-on: http://review.gluster.org/11722
NetBSD-regression: NetBSD Build System 
CentOS-regression: Gluster Build System 
Smoke: Gluster Build System 
Reviewed-by: Atin Mukherjee

features/locks: Implement mandatory locks

2016-05-02T11:18:44+00:00

Initial change to fix/enable the mandatory locking support in GlusterFS
as per the following design:

https://review.gluster.org/#/c/12014/

Accordingly 'locks.mandatory-locking' option is available as part of this
change which will accept one among the following values:

* off
* file
* forced
* optimal

See design doc for more details

Change-Id: I14c489b3f8af5ebcbfa155a03f0c175e9558ac46
BUG: 762184
Signed-off-by: Anoop C S 
Reviewed-on: http://review.gluster.org/9768
Smoke: Gluster Build System 
CentOS-regression: Gluster Build System 
NetBSD-regression: NetBSD Build System 
Reviewed-by: Poornima G 
Reviewed-by: Raghavendra Talur 
Reviewed-by: Rajesh Joseph 
Reviewed-by: Pranith Kumar Karampuri

WORM/Retention Translator: Implementation of file level WORM

2016-05-02T01:05:22+00:00

To activate the file level worm feature, the features.read-only and
features.worm options should be switched "off" on the volume and
the features.worm-file-level should be switched "on". Both read-only
and worm or worm-file-level cannot be switched "on" together. The
files which are created when the worm-file-level option is set on the
volume will have their own retention profile.

If both worm and worm-file-level are "on" at that time the worm
which is the volume level worm will have priority over file level
worm. If worm-file level is switched "off" after some time and the
read-only option is switched "on" then read-only will have priority.

The current implementation allows the users to manually transmit
a file to a WORM-Retained state by removing all the write bits of
the file using the chmod command. The file will have a retention
profile which contains the state of the file, mode of retention,
and the default retention time.

The file will be made WORM-Retained for a default of 120 seconds
during which it will be immutable and undeletable and it sets the
atime of the file to the time till which it is retained.
After that period if any fop request comes for that file, will
make the transition from WORM-Retained state to WORM state, where
the file will be immutable but deletable and, it will reset
the atime to the actual atime of the file. If a WORM file needs
to be made undeletable again, it can be done by using the chmod
command with all the write bits removed.

There are two modes of retention:
1. Relax: where the retention time of a WORM-Retained file can be
   increased or decreased.
2. Enterprise: where the retention time of a WORM-Retained file
   can be increased but not be decreased.
Whenever a utime change(touch -a, -t, ...)request comes for a
file it checks the mode of retention before setting the utimes.
This is done only if the file is WORM-Retained but for a WORM file
it will change the utimes.

Lazy auto commit:
Whenever a file gets created it will store the creation time of the
file or if a file already exists then any of the next unlink, link,
truncate or rename fops will set the current time as the start time
in an xattr. The next rename/unlink/truncate/link call will check for the
auto commit period and if is is expired, then it will automatically do
the state transition. If it is a normal file then it gets converted
to WORM-Retained state. If it is a WORM-Retained file and its retention
period is expired, then it gets converted to WORM state.

Added the volume set options for the WORM translator. It allows the users
to change the default values of auto-commit-period, default-retention-period,
retention-mode. To make use of the file-level WORM first we have to set the
'worm-file' option to 'on'. The files which are created when the worm-file
option is set on the volume will get WORM-Retained. Other files will work
as usual and will not be WORMed. The auto-commit-period, retention-mode,
and the default-retention-period values for the file will be set to the values
which are set on the volume when the file is created.

Added the tests to check the basic functionalities of the WORM/Retention feature.

Change-Id: I77bd9777f9395a944d76b5cc35a5b48a3c14d148
BUG: 1326308
Signed-off-by: karthik-us 
Reviewed-on: http://review.gluster.org/13429
Reviewed-by: Niels de Vos 
Smoke: Gluster Build System 
NetBSD-regression: NetBSD Build System 
CentOS-regression: Gluster Build System

glusterd: volume set changes for lock migration

2016-05-02T01:05:03+00:00

Change-Id: I48c6f9cdda47503615ba65882acd5eedf0a70c89
BUG: 1326085
Signed-off-by: Susant Palai 
Reviewed-on: http://review.gluster.org/14024
Smoke: Gluster Build System 
Reviewed-by: Niels de Vos 
NetBSD-regression: NetBSD Build System 
CentOS-regression: Gluster Build System

features/trash: wind mkdir with special pid

2016-04-27T21:24:48+00:00

Recent changes done w.r.t handling of mkdir calls in posix translator
resulted in crashing the brick process from trash translator. This was
due to the changes made in posix translator to return EPERM for every
mkdir calls without 'gfid-req' set in dictionary. In order to avoid
gfid mismatches during directory creation from brick side trash
translator does not set 'gfid-req'. This patch is to have an exemption
for trash based on a special pid set for those mkdir calls originating
from trash translator and to reset it in callback.

This patch also includes a small optimization to the existing test case
for trash feature.

Change-Id: I59f084ac875e54342ecf2bffa6e43ebd84814153
BUG: 1317361
Signed-off-by: Anoop C S 
Reviewed-on: http://review.gluster.org/13776
Smoke: Gluster Build System 
NetBSD-regression: NetBSD Build System 
CentOS-regression: Gluster Build System 
Reviewed-by: Jeff Darcy