| Commit message (Collapse) | Author | Age | Files | Lines |
... | |
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
This test script verifies below scenarios:
1)Sub-volume is down copy directory
2)Sub-volume is down copy directory - destination dir hash to
up sub-volume
3)Sub-volume is down copy newly created directory - destination
dir hash to up sub-volume
4)Sub-volume is down copy newly created directory - destination
dir hash to down sub-volume
Change-Id: I22b9bf79ef4775b1128477fb858c509a719efb4a
Signed-off-by: sayaleeraut <saraut@redhat.com>
|
|
|
|
|
|
|
|
|
|
|
| |
Three Scenarios:
- Simulate gfid split brain files under a directory
- Resolve gfid splits using `source-brick`, `bigger-file` and
`latest-mtime` methods
- Validate all the files are healed and data is consistent
Change-Id: I8b143f341c0db2f32086ecb6878cbfe3bdb247ce
Signed-off-by: Leela Venkaiah G <lgangava@redhat.com>
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Test case:
1. Create, start and mount a volume.
2. Create a directory on the mount point and start
linux utar.
3. Create another directory on the mount point and
start rsync of linux untar directory.
4. Add bricks to the volume
5. Trigger rebalance on the volume.
6. Wait for rebalance to complete on volume.
7. Wait for I/O to complete.
8. Validate if checksum of both the untar and rsync is same.
Change-Id: I008c65b1783d581129b4c35f3ff90642fffe29d8
Signed-off-by: kshithijiyer <kshithij.ki@gmail.com>
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Steps-
1. Create a volume and mount it
2. Start gluster compilation
3. Bring down redundant bricks
4. Wait for compilation to complete
5. Bring up bricks
6. Check if mountpoint is accessible
7. Delete glusterfs from mountpoint and
start gluster compilation again
8. Bring down redundant bricks
9. Wait for compilation to complete
10. Bring up bricks
11. Check if mountpoint is accessible
Change-Id: Ic5a272fba7db9707c4acf776d5a505a31a34b915
Signed-off-by: ubansal <ubansal@redhat.com>
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Problem:
The code fails if we have give hostname in glusto-tests
config file. This is becuase we have a converstion logic
present in the testcase which converts IP to hostname.
Solution:
Adding code to check if it's an IP and then only
run the code to convert it.
Change-Id: I3bb1a566d469a4c32161c91fa610da378d46e77e
Signed-off-by: kshithijiyer <kshithij.ki@gmail.com>
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Adding the following testcases for block, character
and pipe files:
Test case 1:
1. Create distributed volume with 5 sub-volumes,
start and mount it.
2. Create character and block device files.
3. Check filetype of files from mount point.
4. Verify that the files are stored on only the bricks which is
mentioned in trusted.glusterfs.pathinfo xattr.
5. Verify stat output from mount point and bricks.
Test case 2:
1. Create distributed volume with 5 sub-volumes,
start and mount it.
2. Create character and block device files.
3. Check filetype of files from mount point.
4. Verify that the files are stored on only one bricks which is
mentioned in trusted.glusterfs.pathinfo xattr.
5. Delete the files.
6. Verify if the files are delete from all the bricks
Test case 3:
1. Create distributed volume with 5 sub-volumes,
start and mount it.
2. Create character and block device files.
3. Check filetype of files from mount point.
4. Set a custom xattr for files.
5. Verify that xattr for files is displayed on mount point and bricks.
6. Modify custom xattr value and verify that xattr for files
is displayed on mount point and bricks.
7. Remove the xattr and verify that custom xattr is not displayed.
8. Verify that mount point and brick shows pathinfo xattr properly.
Test case 4:
1. Create distributed volume with 5 sub-volumes,
start and mount it.
2. Create a pipe file.
3. Check filetype of files from mount point.
4. Verify that the files are stored on only the bricks which is
mentioned in trusted.glusterfs.pathinfo xattr.
5. Verify stat output from mount point and bricks.
6. Write data to fifo file and read data from fifo file
from the other instance of the same client.
Upstream bug: https://github.com/gluster/glusterfs/issues/1461
Change-Id: I0e72246ba3d6d20a5de95a95d51271337b6b5a57
Signed-off-by: kshithijiyer <kshithij.ki@gmail.com>
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Steps:
- Create and mount a replicated volume and disable quorum,
self-heal deamon
- Create ~10 files from the mount point and simulate data, metadata
split-brain for 2 files each
- Create a dir with some files and simulate entry/gfid split brain
- Validate volume successfully recognizing split-brain
- Validate a lookup on split-brain files fails with EIO error on mount
- Validate `heal info` and `heal info split-brain` command shows only
the files that are in split-brain
- Validate new files and dir's can be created from the mount
Change-Id: I8caeb284c53304a74473815ae5181213c710b085
Signed-off-by: Leela Venkaiah G <lgangava@redhat.com>
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Problem:
brickdir.hashrange_contains_hash() returns true
or False. However it test test_create_file it's
check it ret == 1 or not
Fix:
Changing ret == 1 to ret.
Change-Id: I53655794f10fc5d778790bdffbe65563907bef6d
Signed-off-by: kshithijiyer <kshithij.ki@gmail.com>
|
|
|
|
|
|
|
|
| |
- `str.rsplit` doesn't accept named args in py2
- Removed named arg to make it compatible with both versions
Change-Id: Iba287ef4c98ebcbafe55f2166c99aef0c20ed9aa
Signed-off-by: Leela Venkaiah G <lgangava@redhat.com>
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Steps:
1.Create a volume and mount it
2.Create deep directory and file in each directory
3.Rename the file
4.Check if brickpath contains old files
5.Delete all data
6.Check .glusterfs/indices/xattrop is empty
7.Check if brickpath is empty
Change-Id: I04e50ef94379daa344be1ae1d19cf2d66f8f460b
Signed-off-by: ubansal <ubansal@redhat.com>
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Steps:
1. Create a volume and mount it
2. Disable heal and cluster-quorum-count
3. Bring down one data and arbiter brick from one
subvol
4. Write IO and validate it
5. Bring up bricks
6. Bring down another data brick and arbiter brick
from the same subvol
7. Write IO and validate it
8. Bring up bricks
9. Check if split-brain is created
10. Write IO -> should fail
11. Enable heal and cluster-quorum-count
12. Write IO -> should fail
Change-Id: I229b58c1bcd70dcd87d35dc410e12f51b032b9c4
Signed-off-by: ubansal <ubansal@redhat.com>
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Test Steps:
1. Create a distributed-replicated(3X3)/distributed-arbiter(3X(2+1))
and mount it on one client
2. Kill 3 bricks corresponding to the 1st subvol
3. Unmount and remount the volume on the same client
4. Create deep dir from mount point 'dir1/subdir1/deepdir1'
5. Create files under dir1/subdir1/deepdir1; touch <filename>
6. Now bring all sub-vols up by volume start force
7. Validate backend bricks for dir creation, the subvol which is
offline will have no dirs created, whereas other subvols will have
dirs created from step 4
8. Trigger heal from client by '#find . | xargs stat'
9. Verify that the directory entries are created on all back-end bricks
10. Create new dir (dir2) on location dir1/subdir1/deepdir1
11. Trigger rebalance and wait for the completion
12. Check backend bricks for all entries of dirs
Change-Id: I4d8f39e69c84c28ec238ea73935cd7ca0288bffc
Signed-off-by: Bala Konda Reddy M <bala12352@gmail.com>
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Problem:
In most of the testcases due to redundant logging,
the performance of the whole suite completion time
is affected.
Solution:
In BVT Test suite, there are 184 g.log.info messages
more than half of them are redundant.
Removed logs wherever it is not required.
Added missing get_super_method for setUp and tearDown
for one testcase, modified increment in the test.
Change-Id: I19e4462f2565906710c2be117bc7c16c121ddd32
Signed-off-by: Bala Konda Reddy M <bala12352@gmail.com>
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Test Steps:
1) Create 1x3 volume and fuse mount the volume
2) On mount created a dir dir1
3) Pkill glusterfsd on node n1 (b2 on node2 and b3 and node3 up)
4) touch f{1..10} on the mountpoint
5) b2 and b3 xattrs would be blaming b1 as files are created while
b1 is down
6) Reset the b3 xattrs to NOT blame b1 by using setattr
7) Now pkill glusterfsd of b2 on node2
8) Restart glusterd on node1 to bring up b1
9) Now bricks b1 online , b2 down, b3 online
10) touch x{1..10} under dir1 itself
11) Again reset xattr on node3 of b3 so that it doesn't blame b2,
as done for b1 in step 6
12) Do restart glusterd on node2 hosting b2 to bring all bricks online
13) Check for heal info, split-brain and arequal for the bricks
Change-Id: Ieea875dd7243c7f8d2c6959aebde220508134d7a
Signed-off-by: Bala Konda Reddy M <bala12352@gmail.com>
|
|
|
|
|
|
|
|
| |
Add logic to do ls -l before and after.
Add logic to set all log-levels to debug.
Change-Id: I512e3b229fe9e2126f6c596fdc031c00a25fbe0b
Signed-off-by: kshithijiyer <kshithij.ki@gmail.com>
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Steps:
1.Create a volume and mount it
2.Starting writing and reading data on file
3.Bring down 1 brick
4.Validate read and write to file
5.Bring up brick and start healing
6.Monitor healing and completion
7.Bring down 2nd brick
8.Read and write to same file
9.Bring up brick and start healing
10.Monitor healing and completion
11.Check split-brain
Change-Id: Ib03a1ad7ee626337904b084e85eee38750fea141
Signed-off-by: ubansal <ubansal@redhat.com>
|
|
|
|
|
|
|
|
|
|
| |
- Validate `heal info` returns before timeout with IO
- Validate `heal info` returns before timeout with IO and brick down
- Validate data heal on file append in AFR, arbiter
- Validate entry heal on file append in AFR, arbiter
Change-Id: I803b931cd82d97b5c20bd23cd5670cb9e6f04176
Signed-off-by: Leela Venkaiah G <lgangava@redhat.com>
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Problem:
In most of the testcases due to redundant logging,
the performance of the whole suite completion time
is affected.
Solution:
Currently there are 100+ g.log.info statements in the
authentincation suite and half of them are redundant.
Removed the g.log.info statements whereever it is not
required. After the changes the g.log.info statements
are around 50 and not removed the statements to reduce
the number of lines but for the improvement of the
whole suite.
Modified few line indents as well and added teardown
for the missing files.
Note: Will be submitting for each components separately
Change-Id: I63973e115dd5dbbc7fc9462978397e7915181265
Signed-off-by: Bala Konda Reddy M <bala12352@gmail.com>
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Problem:
Due to patch [1] which was sent for issue #24
causes a large number of testcases to fail
or get stuck in the latest DHT run.
Solution:
Make changes sot that getfattr command
sends back the output in text wherever needed.
Links:
[1] https://review.gluster.org/#/c/glusto-tests/+/24841/
Change-Id: I6390e38130b0699ceae652dee8c3b2db2ef3f379
Signed-off-by: kshithijiyer <kshithij.ki@gmail.com>
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Problem:
Testcases test_volume_start_stop_while_rebalance_is_in_progress throws
the below traceback when run:
```
Traceback (most recent call last):
File "/usr/lib64/python2.7/logging/__init__.py", line 851, in emit
msg = self.format(record)
File "/usr/lib64/python2.7/logging/__init__.py", line 724, in format
return fmt.format(record)
File "/usr/lib64/python2.7/logging/__init__.py", line 464, in format
record.message = record.getMessage()
File "/usr/lib64/python2.7/logging/__init__.py", line 328, in getMessage
msg = msg % self.args
TypeError: not all arguments converted during string formatting
Logged from file test_volume_start_stop_while_rebalance_in_progress.py, line 135
```
This is because g.log.error() was used instead of
self.assertTrue().
Solution:
Changing to self.assertTrue().
Change-Id: If926eb834c0128a4e507da9fdd805916196432cb
Signed-off-by: kshithijiyer <kshithij.ki@gmail.com>
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
1. Create volume and create files/dirs from mount point
2. With IO in progress execute reset-brick start
3. Now format the disk from back-end, using rm -rf <brick path>
4. Execute reset brick commit and check for the brick is online.
5. Issue volume heal using "gluster vol heal <volname> full"
6. Check arequal for all bricks to verify all backend bricks
including the resetted brick have same data
Change-Id: I06b93d79200decb25f863e7a3f72fc8e8b1c4ab4
Signed-off-by: Bala Konda Reddy M <bala12352@gmail.com>
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Test Steps:
1. Create a pure-ec volume (say 1x(4+2))
2. Mount volume on two clients
3. Create some files and dirs from both mnts
4. Add bricks in this case the (4+2) ie 6 bricks
5. Create a new dir(common_dir) and in that directory create a distinct
directory(using hostname as dirname) for each client and pump IOs
from the clients(dd)
6. While IOs are in progress replace any of the bricks
7. Check for errors if any collected after step 6
Change-Id: I3125fc5906b5d5e0bc40477e1ed88825f53fa758
Signed-off-by: Bala Konda Reddy M <bala12352@gmail.com>
|
|
|
|
|
|
|
|
| |
TC's was failing due to timeout issue
increased reabalnce timeout from 900 to 1800
Change-Id: I726217a21ebbde6391660dd3c9dc096cc9ca6bb4
Signed-off-by: ubansal <ubansal@redhat.com>
|
|
|
|
|
| |
Change-Id: I813f3e78ad8b0b79940635df6721e34e6bc93f34
Signed-off-by: Manisha Saini <msaini@redhat.com>
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Steps:
1.Create a volume and mount it
2.Create a directory say d1
3.Create deep directories and files in d1
4.Bring down redundant bricks
5.Delete d1
6.Create d1 and same data again
7.Bring bricks up
8.Monitor heal
9.Verify split-brain
Change-Id: I778fab6bf6d9f81fca79fe18285073e1f7ccc7e7
Signed-off-by: ubansal <ubansal@redhat.com>
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Problem:
Test script test_alert_time_out currently fails
2 out of 6 times when executed on the same setup
this is due to the log files not have 120004 A
alert message. This issue is only observed in
distributed volume type mounted over fuse
protocol.
Solution:
There is no permanent solution to this problem
as even if we increase the sleep 20 seconds there
is still a chance that it might fail. The optimal
sleep time where it only fails 5 times after 15
attempts is 6 seconds. Hence changing sleep time
to 6 seconds.
Change-Id: I9e9bd41321e24f502d90c3c34edce9113133755e
Signed-off-by: kshithijiyer <kshithij.ki@gmail.com>
|
|
|
|
|
|
|
| |
The assertIsNotNone is missing the param.
Change-Id: Iddff9b203672b2edf702ada624bfac1892641712
Signed-off-by: Pranav <prprakas@redhat.com>
|
|
|
|
|
|
|
|
|
| |
Adding code to get dir tree and dump all
xattr in hex for Bug 1810901 before remove-brick
also adding logic to set log-level to debug.
Change-Id: I9c9c970c4de7d313832f6f189cdca8428a073b1e
Signed-off-by: kshithijiyer <kshithij.ki@gmail.com>
|
|
|
|
|
|
|
| |
Integrate the changes made in library to the test
Change-Id: I9bf8c3f1f732132170a96405a4a12839463a2eaa
Signed-off-by: Pranav <prprakas@redhat.com>
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Description: Checks that there is no data loss when remove-brick
operation is stopped and then new bricks are added to
the volume.
Steps:
1) Create a volume.
2) Mount the volume using FUSE.
3) Create files and dirs on the mount-point.
4) Calculate the arequal-checksum on the mount-point.
5) Start remove-brick operation on the volume.
6) While migration is in progress, stop the remove-brick
operation.
7) Add-bricks to the volume and trigger rebalance.
8) Wait for rebalance to complete.
9) Calculate the arequal-checksum on the mount-point.
Change-Id: I96a7311f5acd0ae19b17d7b7c7da4d3899cdef77
Signed-off-by: sayaleeraut <saraut@redhat.com>
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Steps:
- Create a volume and mount it
- disable metadata,data,entry heal
- Create files and take arequal of mount point
- Bring down redundant bricks
- Append data and create hardlinks
- Bring up bricks
- Check healing and split-brain
- Bring down redundant bricks
- Truncate data
- Check file and hardlink stat match
- Bring up bricks
Change-Id: I9b26f2fb26d72b71abd63a25ef8d9173f32997d4
Signed-off-by: ubansal <ubansal@redhat.com>
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Problem:
Testcases test_mount_snap_delete and test_restore_online_vol
were failing in the latest runs with the below traceback
```
Traceback (most recent call last):
File "/usr/share/glustolibs/io/scripts/file_dir_ops.py", line 1246, in <module>
rc = args.func(args)
File "/usr/share/glustolibs/io/scripts/file_dir_ops.py", line 374, in create_files
base_file_name, file_types)
File "/usr/share/glustolibs/io/scripts/file_dir_ops.py", line 341, in _create_files
ret = pool.map(lambda file_tuple: _create_file(*file_tuple), files)
File "/usr/lib64/python3.6/multiprocessing/pool.py", line 266, in map
return self._map_async(func, iterable, mapstar, chunksize).get()
File "/usr/lib64/python3.6/multiprocessing/pool.py", line 644, in get
raise self._value
File "/usr/lib64/python3.6/multiprocessing/pool.py", line 119, in worker
result = (True, func(*args, **kwds))
File "/usr/lib64/python3.6/multiprocessing/pool.py", line 44, in mapstar
return list(map(*args))
File "/usr/share/glustolibs/io/scripts/file_dir_ops.py", line 341, in <lambda>
ret = pool.map(lambda file_tuple: _create_file(*file_tuple), files)
File "/usr/share/glustolibs/io/scripts/file_dir_ops.py", line 270, in _create_file
with open(file_abs_path, "w+") as new_file:
FileExistsError: [Errno 17] File exists: '/mnt/testvol_distributed_glusterfs/file1.txt'
```
This was because I/O logic was trying to create 2 files with the same
name from 2 clients.
Fix:
Modify logic to use counters to create files with different names.
Change-Id: I2896736d28f6bd17435f941088fd634347e3f4fd
Signed-off-by: kshithijiyer <kshithij.ki@gmail.com>
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Steps:
- Create an volume and mount it
- Bring bricks offline
- Write 50k files
- Bring bricks online
- Monitor heal completion
- Check for split-brain
Change-Id: I40739effdfa1c1068fa0628467154b9a667161a3
Signed-off-by: ubansal <ubansal@redhat.com>
|
|
|
|
|
|
|
|
| |
Adding code to get dir tree and dump all
xattr for Bug 1810901.
Change-Id: Ia59dcd2623e845066e31037c96a64249efa074c2
Signed-off-by: kshithijiyer <kshithij.ki@gmail.com>
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Steps:
- Create, start and mount an arbiter volume in two clients
- Create two dir's, fill IO in first dir and take note of arequal
- Start a continuous IO from second directory
- Convert arbiter to x2 replicated volume (remove brick)
- Convert x2 replicated to x3 replicated volume (add brick)
- Wait for ~5 min for vol file to be updated on all clients
- Enable client side heal options and issue volume heal
- Validate heal completes with no errors and arequal of first dir
matches against initial checksum
Change-Id: I291acf892b72bc8a05e76d0cffde44d517d05f06
Signed-off-by: Leela Venkaiah G <lgangava@redhat.com>
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Steps:
- Create and mount a replicated volume
- Kill one of the bricks and write IO from mount point
- Verify `gluster volume heal <volname> info healed` and `gluster
volume heal <volname> info heal-failed` command results in error
- Validate `gluster volume help` doesn't list `healed` and
`heal-failed` commands
Change-Id: Ie1c3db12cdfbd54914e61f812cbdac382c9c723e
Signed-off-by: Leela Venkaiah G <lgangava@redhat.com>
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Problem:
NFS-Ganesha Tests inherits 'NfsGaneshaClusterSetupClass' whereas
the other tests inherits 'GlusterBaseClass'. This causes a cyclic
dependency when trying to run other modules with Nfs-Ganesha.
Fix:
1. Move the Nfs-Ganesha dependencies to GlusterBaseClass
2. Modify the Nfs-Ganesha tests to inherit from GlusterBaseClass
3. Remove setup_nfs_ganesha method call from existing Ganesha tests
as its invoked by default from GlusterBaseClass.SetUpClass
Change-Id: I1e382fdb2b29585c097dfd0fea0b45edafb6442b
Signed-off-by: Pranav <prprakas@redhat.com>
|
|
|
|
|
| |
Change-Id: I16c5f070d807673662e5ac3583aace06873a9c14
Signed-off-by: Manisha Saini <msaini@redhat.com>
|
|
|
|
|
|
|
|
|
|
| |
Description:
Sos must be able to capture the required logs in sosreport including
gluster logs, without compromising the integrity of Gluster like
deleting socket files etc
Change-Id: Ifec57778ff5d1fc0ceaa3ecf94a9851244076d2b
Signed-off-by: nchilaka <nchilaka@redhat.com>
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Problem: Test is failing with below traceback
when ran with python3 as default.
`
Traceback (most recent call last):
File "<string>", line 1, in <module>
TypeError: a bytes-like object is required, not 'str'
`
Solution:
Added ''.encode() which will fix the issue when ran
using both python2 and python3
Added a check for core file on the client node.
Change-Id: I8f800f5fad97c3b7591db79ea51203e5293a1f69
Signed-off-by: Bala Konda Reddy M <bala12352@gmail.com>
|
|
|
|
|
|
|
|
|
|
| |
- Remove unneccessary disablement of client side heal options
- Check if client side heal options are disabled by default
- Test data heal by default method
- Explicit data heal by calling self heal command
Change-Id: I3be9001fc1cf124a4cf5a290cee985e166c0b685
Signed-off-by: nchilaka <nchilaka@redhat.com>
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Description : Check that all directories are read and listed while
rebalance is still in progress.
Steps :
1) Create a volume.
2) Mount the volume using FUSE.
3) Create a dir "master" on mount-point.
4) Create 8000 empty dirs (dir1 to dir8000) inside dir "master".
5) Now inside a few dirs (e.g. dir1 to dir10), create deep dirs
and inside every dir, create 50 files.
6) Collect the number of dirs present on /mnt/<volname>/master
7) Change the rebalance throttle to lazy.
8) Add-brick to the volume (at least 3 replica sets.)
9) Start rebalance using "force" option on the volume.
10) List the directories on dir "master".
Change-Id: I4d04b3e2be93b5c25b5ed70516bb99d99fb1fb8a
Signed-off-by: sayaleeraut <saraut@redhat.com>
|
|
|
|
|
|
|
|
|
| |
- `replace` funciton to used to forgo version check
- `unicode` is not being recognized from builtins in py2
- `replace` seems correct alternative than fixing unicode
Change-Id: Ieb9b5ad283e1a31d65bd8a9715b80f9deb0c05fe
Signed-off-by: Leela Venkaiah G <lgangava@redhat.com>
|
|
|
|
|
|
|
|
|
| |
- Remove decimal before passing to `head` command
- Breakup sparsefile into chunks to ~half of brick size
- Whole test has to be skipped due to BZ #1339144
Change-Id: I7a9ae25798b442c74248954023dd821c3442f8f9
Signed-off-by: Leela Venkaiah G <lgangava@redhat.com>
|
|
|
|
|
|
|
|
|
|
|
|
| |
Problem: Creating a third mount obj works for glusterfs
protocol but in future while running for nfs/cifs might
face complications and test might fail.
Solution: Skip test unless three clients are provided
Removing redundant logging and minor fixes.
Change-Id: Ie657975a46b6989cb9f057f5cc337333bbf1010d
Signed-off-by: Bala Konda Reddy M <bala12352@gmail.com>
|
|
|
|
|
|
|
| |
- Translate function is availble on `unicode` string in Python2
Change-Id: I6aa01606acc73b18d889a965f1c01f9a393c2c46
Signed-off-by: Leela Venkaiah G <lgangava@redhat.com>
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Test steps:
1. Create a volume, start and mount it on a client
2. Bring down redundant bricks in the subvol
3. Create a file on the volume using "touch"
4. Truncate the file using "O_TRUNC"
5. Bring the brick online
6. Write data on the file and wait for heal completion
7. Check for crashes and coredumps
Change-Id: Ie02a56ab5180f6a88e4499c8cf6e5fe5019e8df1
Signed-off-by: Bala Konda Reddy M <bala12352@gmail.com>
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Test Steps:
1. Created a volume and mount this volume on 3 clients.
2. Bring down two bricks offline in each subvol.
3. On client1: under dir1 create files f{1..10000} run in background
4. On client2: under / touch x{1..1000}
5. On client3: start creating x{1001..10000}
6. Bring bricks online which were offline(brought up all the bricks
which were down (2 in each of the two subvols)
7. While IO on Client1 and Client3 were happening, On client2 move
all the x* files into dir1
8. Perform lookup from client 3
Change-Id: Ib72648af783535557e20cea7e64ea68036b23121
Signed-off-by: Bala Konda Reddy M <bala12352@gmail.com>
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Steps:
1.Create a EC volume and mount it
2.Run different types of IO's
3.Take arequal of mountpoint
4.Bring down redundant bricks
5.Take arequal of mountpoint
6.Bring down another set of redundant bricks
7.Take arequal of mountpoint
Change-Id: If253cdfe462c6671488e858871ec904fbb2f9ead
Signed-off-by: ubansal <ubansal@redhat.com>
|
|
|
|
|
| |
Change-Id: I8ae78b06706bc4818cbd2b00b386f362883cb9d7
Signed-off-by: Manisha Saini <msaini@redhat.com>
|