summaryrefslogtreecommitdiffstats
path: root/tests
Commit message (Collapse)AuthorAgeFilesLines
* [Test]: Validate EC Eagerlock Behavior and Performancenchilaka2020-06-191-0/+264
| | | | | | | | | | Description: This script tests Disperse(EC) eagerlock default values and the performance impact on lookups with eagerlock and other-eagerlock default values Change-Id: Ia083d0d00f99a42865fb6f06eda75ecb18ff474f Signed-off-by: nchilaka <nchilaka@redhat.com>
* [Test] Add tc to check eager lock clikshithijiyer2020-06-191-0/+71
| | | | | | | | | | | | Testcase Steps: 1.Create an EC volume 2.Set the eager lock option by turning on disperse.eager-lock by using different inputs: - Try non boolean values(Must fail) - Try boolean values Change-Id: Iec875ce9fb4c8f7c68b012ede98bd94b82d04d7e Signed-off-by: kshithijiyer <kshithij.ki@gmail.com>
* [Test] Validate delete file picked for migrationsayaleeraut2020-06-191-0/+165
| | | | | | | | | | | | | | | | | | | Description: The test script verifies that if a file is picked for migration and if it is deleted, then the file should be removed successfully. Steps : 1) First create a big data file of 10GB. 2) Rename that file, such that after rename a linkto file is created (we are doing this to make sure that file is picked for migration.) 3) Add bricks to the volume and trigger rebalance using force option. 4) When the file has been picked for migration, delete that file from the mount point. 5) Check whether the file has been deleted or not on the mount-point as well as the back-end bricks. Change-Id: I137512a1d94a89aa811a3a9d61a9fb4002bf26be Signed-off-by: sayaleeraut <saraut@redhat.com>
* [Test] Add dht file rename cases where the destination file existsPranav2020-06-191-0/+540
| | | | | | | | Tests to validate behaviour of different scenarios of file rename cases, when destination file exists intially. Change-Id: I12cd2568540bec198f1c3cf85213e0107c9ddd6b Signed-off-by: Pranav <prprakas@redhat.com>
* [Testfix] Add reset-failed cmdkshithijiyer2020-06-181-3/+7
| | | | | | | | | | | | | | | | | | | | Problem: The testcase test_volume_create_with_glusterd_restarts consist of a asynchronous loop of glusterd restarts which fails in the lastest runs due to patch [1] and [2] added to glusterfs which limits the glusterd restarts to 6. Fix: Add `systemctl reset-failed glusterd` to the asynchronous loop. Links: [1] https://review.gluster.org/#/c/glusterfs/+/23751/ [2] https://review.gluster.org/#/c/glusterfs/+/23970/ Change-Id: Idd52bfeb99c0c43afa45403d71852f5f7b4514fa Signed-off-by: kshithijiyer <kshithij.ki@gmail.com>
* [TestFix] Remove 'replicated' from volume typeLeela Venkaiah G2020-06-171-1/+1
| | | | | | | - Test is designed to run on EC volumes only Change-Id: Ice6a77422695ebabbec6b9cfd910e453e5b2c81a Signed-off-by: Leela Venkaiah G <lgangava@redhat.com>
* [Test] Checks all types of heal on EC volumeubansal2020-06-171-0/+285
| | | | | | | | | | | | | | | | Steps: 1.Create a volume and mount it 2.Create a directory, dir1 and run different types of IO's 3.Create a directory, dir2 4.Bring down redundant bricks 5.Write IO's to directory dir2 6.Create a directory, dir3 and run IO's(read,write,ammend) 7.Bring up bricks 8.Monitor heal 9.Check for data intergrity of dir1 Change-Id: I9a7e366084bb46dcfc769b1d98b89b303fc16150 Signed-off-by: ubansal <ubansal@redhat.com>
* [Test] Rename of filetypes on brick downLeela Venkaiah G2020-06-171-0/+221
| | | | | | | | | | | | | | | | | | Test Steps: 1. Create an EC volume 2. Mount the volume using FUSE on two different clients 3. Create ~9 files from one of the client 4. Create ~9 dir with ~9 files each from another client 5. Create soft-links, hard-links for file{4..6}, file{7..9} 6. Create soft-links for dir{4..6} 7. Begin renaming the files, in multiple iterations 8. Bring down a brick while renaming the files 9. Bring the brick online after renaming some of the files 10. Wait for renaming of the files 11. Validate no data loss and files are renamed successfully Change-Id: I6d98c00ff510cb473978377bb44221908555681e Signed-off-by: Leela Venkaiah G <lgangava@redhat.com>
* [Test] Git clone on glusterfs volumeBala Konda Reddy M2020-06-161-0/+80
| | | | | | | | | | | Test Steps: 1. Create a volume and mount it on one client 2. git clone the glusterfs repo on the glusterfs volume 3. Set the performance options to off 4. Repeat step 2 on a different directory Change-Id: Iaecce7cd14ecf84058c75847a037c6589d3833e9 Signed-off-by: Bala Konda Reddy M <bala12352@gmail.com>
* [Test]: Check File descriptor in ECubansal2020-06-161-0/+174
| | | | | | | | | | Steps: 1.Open a File descriptor when a brick in down 2.Write to File descriptor when brick has come up and check if healing is complete Change-Id: I721cedf4dc6a420f0c153d4232b046f780da201b Signed-off-by: ubansal <ubansal@redhat.com>
* [Test] Add dht file rename cases where destination doesn't existPranav2020-06-151-0/+450
| | | | | | | | Tests to validate behaviour of different scenarios of file rename cases, when destination file doesn't exist intially. Change-Id: I3f22d61d9bd2fa5c54930715e2ef976c7d1ba54e Signed-off-by: Pranav <prprakas@redhat.com>
* [Testfix] Fix teardown failure for test_deem_statfs_quotad.pyPranav2020-06-101-2/+2
| | | | | | | | | | | | | | | Problem: As in setUpClass, the setup_volume method is used for volume creation, volume mount is not done. But in teardown, unmount_volume_and_cleanup_volume is used. This causes teardown to fail Fix: change unmount_volume_and_cleanup_volume to cleanup_volume Change-Id: Ia9bb6bcd36ce9ddb9c200ef18779df47a009d42f Signed-off-by: Pranav <prprakas@redhat.com>
* [TestFix] Decreased the IO's and increase timeoutsubansal2020-06-101-4/+6
| | | | | | | | | | | Decreased the deep directory creation from 15 to 10 As heal completion within 20 mins for disperse was intermitently failing Also increased timeout for bricks to be online and healing to complete Change-Id: I1c1eef383ca4bf3f7f1f89e00da096bbbf57b9db Signed-off-by: ubansal <ubansal@redhat.com>
* [Test] Add tc to check max-bricks-per-processkshithijiyer2020-06-101-3/+98
| | | | | | | | | | | | | | | | | | | | | Test Case: - Create a gluster cluster - With brick mux set to disable: 1.Set cluster.max-bricks-per-process to int and check error message(Must fail) 2.Set cluster.max-bricks-per-process to string(Must fail) - With brick mux set to enable: 1.Set cluster.max-bricks-per-process to string(Must fail) 2.Set cluster.max-bricks-per-process to 0 3.Set cluster.max-bricks-per-process to 1 and check error message.(Must fail) 4.Set cluster.max-bricks-per-process to int value > 1. Also fixing small issues observed when running all the tests in the file. Change-Id: Iad27cd5bbeccc2bd2f0a7e510f881b0ffcb0d3b6 Signed-off-by: kshithijiyer <kshithij.ki@gmail.com>
* [LibFix] Correct the name for dist-arbiter volumesayaleeraut2020-06-091-2/+2
| | | | | | | | | | | The volume distributed-arbiter was mentined as "distributed-arbiter: &distrbuted_arbiter" and "type: distributed_arbiter" under "volume_types:"; but it must be "&distrbuted-arbiter" and "type: distributed-arbiter", in accordance with the format in gluster_base_class.py Change-Id: I4a35a8827f26050328acfeddc3ce930225181e7a Signed-off-by: sayaleeraut <saraut@redhat.com>
* [TestFix] Correct the library name for a functionsayaleeraut2020-06-081-3/+3
| | | | | | | | | The function "set_volume_options()" is a part of "volume_ops" lib, but was wrongly imported from "volume_libs" lib earlier. Corrected the import statement. Change-Id: I7295684e7a564468ac42bbe1f00643ee150f769d Signed-off-by: sayaleeraut <saraut@redhat.com>
* [Testfix] Move teardownclass to teardownkshithijiyer2020-06-084-43/+43
| | | | | | | | | | | | | | | | | | Problem: A pattern was observed where testcases which were passing were throwing error in teadownclass this was becuase docleanup was running before teadownclass and when teardownclass was executed it failed as the setup was already cleaned. Solution: Change code to teardown from teardownclass and move setup volume code to setup from setupclass. Change-Id: I37c6fde1f592224c114148f0ed7215b2494b4502 Signed-off-by: kshithijiyer <kshithij.ki@gmail.com>
* [Test] Check stack overflow in readdirpsayaleeraut2020-06-051-0/+131
| | | | | | | | | | | | | | | | | | | Description: The TC checks that there is no stack overflow in readdirp with parallel-readdir enabled. Steps : 1) Create a volume. 2) Mount the volume using FUSE. 3) Enable performance.parallel-readdir and performance.readdir-ahead on the volume. 4) Create 10000 files on the mount point. 5) Add-brick to the volume. 6) Perform fix-layout on the volume (not rebalance). 7) From client node, rename all the files, this will result in creation of linkto files on the newly added brick. 8) Do ls -l (lookup) on the mount-point. Change-Id: I821efea55e3f8981ecd307b93781f2da5790c548 Signed-off-by: sayaleeraut <saraut@redhat.com>
* [Test] Validate Consistent Time(ctime) featurenchilaka2020-06-051-0/+205
| | | | | | | | | Description: This testcase tests the new xattr glusterfs.mdata attributes when features.ctime is enabled Change-Id: I3c2dd6537f88b6d9da85aa6819c96e1c236a0d61 Signed-off-by: nchilaka <nchilaka@redhat.com>
* [Test] Add tc to check brickmux with wrong valueskshithijiyer2020-06-051-0/+30
| | | | | | | | | | | | Test Case: - Create a gluster cluster - Set cluster.brick-multiplex value to random string(Must fail) - Set cluster.brick-multiplex value to random int(Must fail) - Set cluster.brick-multiplex value to random special characters(Must fail) Change-Id: Ib0233668aad8d72572b1dd9d17a5d0c27c364250 Signed-off-by: kshithijiyer <kshithij.ki@gmail.com>
* [Test] Verify heal on EC volume on file appendsLeela Venkaiah G2020-06-051-0/+186
| | | | | | | | | | | | | | Test Steps: - create and mount EC volume 4+2 - start append to a file from client - bring down one of the bricks (say b1) - wait for ~minute and bring down another brick (say b2) - after ~minute bring up first brick (b1) - check the xattrs 'ec.size', 'ec.version' - xattrs of online bricks should be same as an indication to heal Change-Id: I81a5bad4a91dd891fbbc9d93ae3f76610237789e Signed-off-by: Leela Venkaiah G <lgangava@redhat.com>
* [Test] Validates data deletion on EC volumeubansal2020-06-041-0/+270
| | | | | | | | | | | | | | Test Steps: 1. Create a volume, start and mount it 2. Create directories and files 3. Rename, change permissions of files 4. Create hardlink and soflink and different types of IO's 5. Delete all the data 6. Check no heals are pending 7. Check al bricks are empty Change-Id: Ic8f5dad1a44de71688a6b0a2fcfb4a25cef435ba Signed-off-by: ubansal <ubansal@redhat.com>
* [Test] Read from hardlink in disperseBala Konda Reddy M2020-06-031-0/+111
| | | | | | | | | | | | Test steps: 1. Create volume, start and mount it one client. 2. Enable metadata-cache(md-cache) options on the volume. 3. Touch a file and create a hardlink for it. 4. Read data from the hardlink. 5. Read data from the actual file. Change-Id: Ibf4b8757262707fcfb4d09b4b031ff9dea166570 Signed-off-by: Bala Konda Reddy M <bala12352@gmail.com>
* [Test] Validate file rename with brick downsayaleeraut2020-06-031-0/+172
| | | | | | | | | | | | | | | | | | | | | | | | | | Description: The TC checks that there is no data loss when rename is performed with a brick of volume down. Steps : 1) Create a volume. 2) Mount the volume using FUSE. 3) Create 1000 files on the mount point. 4) Create the soft-link for file{1..100} 5) Create the hard-link for file{101..200} 6) Check for the file count on the mount point. 7) Begin renaming the files, in multiple iterations. 8) Let few iterations of the rename complete successfully. 9) Then while rename is still in progress, kill a brick part of the volume. 10) Let the brick be down for sometime, such that the a couple of rename iterations are completed. 11) Bring the brick back online. 12) Wait for the IO to complete. 13) Check if there is any data loss. 14) Check if all the files are renamed properly. Change-Id: I7b7c4aed7df7f19a10ec8c2577dfec1f1ceeb46c Signed-off-by: sayaleeraut <saraut@redhat.com>
* [Test] Check 'storage.reserve' with wrong valuesLeela Venkaiah G2020-05-281-0/+79
| | | | | | | | | | Test Steps: 1) Create and start a distributed-replicated volume. 2) Give different inputs to the storage.reserve volume set options 3) Validate the command behaviour on wrong inputs Change-Id: I4bbad81cbea9b3b9e59a61fcf7f2b70eac19b216 Signed-off-by: Leela Venkaiah G <lgangava@redhat.com>
* [TestFix] Fix assertItemsEqual issue with python3Pranav2020-05-277-26/+36
| | | | | | | | | | | | | | Issue: In python3 assertItemsEqual is no longer supported and is replaced with assertCountEqual (Refer [1]). Because of this issue, few arbiter tests are failing. [1] https://docs.python.org/2/library/unittest.html#unittest.TestCase.assertItemsEqual Fix: The replacement assertCountEqual is not supported in python2. So the fix is to replace assertItemsEqual with assertEqual(sorted(expected), sorted(actual)) Change-Id: Ic1d599fa31f85a8a41598b6c245056a6ff01e000 Signed-off-by: Pranav <prprakas@redhat.com>
* [Test] Verify 'gluster get-state' on brick unmountLeela Venkaiah G2020-05-261-0/+126
| | | | | | | | | | | | | | Testcase steps: 1. Form a gluster cluster by peer probing and create a volume 2. Unmount the brick using which the volume is created 3. Run 'gluster get-state' and validate absence of error 'Failed to get daemon state. Check glusterd log file for more details' 4. Create another volume and start it using different bricks which are not used to create above volume 5. Run 'gluster get-state' and validate the absence of above error. Change-Id: Ib629b53c01860355e5bfafef53dcc3233af071e3 Signed-off-by: Leela Venkaiah G <lgangava@redhat.com>
* [Test] Add tc for root squash on nfs-ganeshaManisha Saini2020-05-262-0/+182
| | | | | | | | | | | | | | | | | | | | Verifcation of root-squash functionality with NFS-Ganesha * Create a volume and export it via Ganesha * Mount the volume on clients * Create some files and dirs inside mount point * Check for owner and group Owner and group should be root * Set permission as 777 for mount point * Enable root-squash on volume * Create some more files and dirs * Check for owner and group for any file Owner and group should be nfsnobody * Edit file created by root user nfsnobody user should not be allowed to edit file Change-Id: Ia345c772c84fcfe6ef716b9f1026fca5d399ab2a Signed-off-by: Manisha Saini <msaini@redhat.com>
* [Test] Quota validation on EC volumeubansal2020-05-211-0/+159
| | | | | Change-Id: I3f77dc73044a5bc59a26319c55e8e024e2edf449 Signed-off-by: ubansal <ubansal@redhat.com>
* [TestFix] DHT Tests - Remove NFS protocolsayaleeraut2020-05-211-10/+11
| | | | | | | | | | | | | | | Scenarios: 1 - Rename directory when destination is not present 2 - Rename directory when destination is present The TC was failing when the volume was mounted using NFS at validate_files_in_dir() because the method uses 'trusted.glusterfs.pathinfo' on the mount, which is a glusterfs specific xattr. When the volume is mounted using NFS, it cannot find the xattr and hence it failed. Change-Id: Ic61de773525e717a73178a4694c015276da2a688 Signed-off-by: sayaleeraut <saraut@redhat.com>
* [Testfix] Fix assertIn in test_enabling_brick_muxkshithijiyer2020-05-201-2/+2
| | | | | | | | | assertIn statement looks for out in warning_message which fails every time as it should ideally look for warning_message in out. Change-Id: I57e0221097c861e251995e5e8456cb19964e7d17 Signed-off-by: kshithijiyer <kshithij.ki@gmail.com>
* [Test] Validate USS after snapshot restoreVinayak Papnoi2020-05-181-0/+239
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | When a snapshot is restored, that snapshot gets removed. USS makes use of a '.snaps' directory in the mount point where all the activated snapshots can be listed. So the restored snapshot should not be listed under the '.snaps' directory regardless of it being activated or deactivated. Steps: * Perform I/O on mounts * Enable USS on volume * Validate USS is enabled * Create a snapshot * Activate the snapshot * Perform some more I/O * Create another snapshot * Activate the second * Restore volume to the second snapshot * From mount point validate under .snaps - first snapshot should be listed - second snapshot should not be listed Change-Id: I5630d8aad6b4758d49e8d4f53497073c78a00a6b Co-authored-by: Sunny Kumar <sunkumar@redhat.com> Signed-off-by: Sunny Kumar <sunkumar@redhat.com> Signed-off-by: Vinayak Papnoi <vpapnoi@redhat.com>
* [Test][Bug] Check for consistent timevalues of a created objectnchilaka2020-05-182-0/+130
| | | | | | | | | | | | | | | Test Summary and Steps: This testcase validates if ctime, mtime and atime of a created object is same 1. Create a volume and check if features.ctime is disabled by default 2. Enable features.ctime 3. Create a new directory dir1 and check if m|a|ctimes are same 4. Create a new file file1 and check if m|a|ctimes are same 5. Again create a new file file2 and check if m|a|ctimes are same after issueing an immediate lookup Change-Id: I024c11706a0309806c081c957b9305be92936f7f Signed-off-by: nchilaka <nchilaka@redhat.com>
* [Test] Add tc to remove brick using nfs_ganeshakshithijiyer2020-05-181-0/+152
| | | | | | | | | | | | | | Verify remove brick operation while IO is running Steps: 1. Start IO on mount points 2. Perform remove brick operation 3. Validate IOs Change-Id: Ie394f96c9180be57704ca637c8cd725af82323cb Co-authored-by: Jilju Joy <jijoy@redhat.com> Signed-off-by: Jilju Joy <jijoy@redhat.com> Signed-off-by: kshithijiyer <kshithij.ki@gmail.com>
* [Testfix] Move test from teardown class to teardownSri Vignesh2020-05-155-66/+29
| | | | | | | Move cases from teardown class to teardown in snapshot Change-Id: I7b33fa2728665fad000a5ad881f6690d40913f22 Signed-off-by: Sri Vignesh <sselvan@redhat.com>
* [Test] test rmdir with subvol downPranav2020-05-131-0/+361
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | test case: (rmdir with subvol down) case -1: - create parent - bring down a non-hashed subvolume for directory child - create parent/child - rmdir /mnt/parent will fail with ENOTCONN case -2: - create dir1 and dir2 - bring down hashed subvol for dir1 - bring down a non-hashed subvol for dir2 - rmdir dir1 should fail with ENOTCONN - rmdir dir2 should fail with ENOTCONN case -3: - create parent - mkdir parent/child - touch parent/child/file - bringdown a subvol where file is not present - rm -rf parent - Only file should be deleted - rm -rf should fail with ENOTCONN case -4: - Bring down a non-hashed subvol for parent_dir - mkdir parent - rmdir parent should fails with ENOTCONN Change-Id: I8fbd425729aaf04eabfced315f94167178918e31 Co-authored-by: Susant Palai <spalai@redhat.com> Signed-off-by: Susant Palai <spalai@redhat.com> Signed-off-by: Pranav <prprakas@redhat.com>
* [Test][BUG] Uss and snapshot in EC volumeubansal2020-05-131-0/+328
| | | | | | | | | | | | | | | Steps: 1.Enable uss and create snapshot, list and delete 2.Create Snapshot with same same and list Github Issue for CentOS-CI failure: https://github.com/gluster/glusterfs/issues/1203 Testcase failing due to: https://bugzilla.redhat.com/show_bug.cgi?id=1828820 Change-Id: I829e6b340dfb4963355b445259fcb011b62ba057 Signed-off-by: ubansal <ubansal@redhat.com>
* [BUG][Test] Add tc to check heal with only shd runningPranav2020-05-111-0/+245
| | | | | | | | | | | | | | | | | | | | | | | | | | | | Failing in CentOS-CI due to this bug https://bugzilla.redhat.com/show_bug.cgi?id=1768380 Description: Test Script which verifies that the server side healing must happen only if the heal daemon is running on the node where source brick resides. * Create and start the Replicate volume * Check the glustershd processes - Only 1 glustershd should be listed * Bring down the bricks without affecting the cluster * Create files on volume * kill the glustershd on node where bricks is running * bring the bricks up which was killed in previous steps * check the heal info - heal info must show pending heal info, heal shouldn't happen since glustershd is down on source node * issue heal * trigger client side heal * heal should complete successfully Change-Id: I1fba01f980a520b607c38d8f3371bcfe086f7783 Co-authored-by: Vijay Avuthu <vavuthu@redhat.com>, Milind Waykole <milindwaykole96@gmail.com> Signed-off-by: Vijay Avuthu <vavuthu@redhat.com> Signed-off-by: Milind Waykole <milindwaykole96@gmail.com> Signed-off-by: Pranav <prprakas@redhat.com>
* [TestFix][Quota] Moving test steps from teardown class to teardownSri Vignesh2020-05-1110-120/+86
| | | | | | | Move cases from teardown class to teardown in quota Change-Id: Ia20fe9bef09842f891f0f27ab711a1ef4c9f6f39 Signed-off-by: Sri Vignesh <sselvan@redhat.com>
* [TestFix] Reduce the workload as rebalance was failingubansal2020-05-111-5/+5
| | | | | Change-Id: Id94870735b26fbeab2bf448d4f80341c92beb5ba Signed-off-by: ubansal <ubansal@redhat.com>
* [TestFix][DHT] Moving test steps from teardown class to teardownSri Vignesh2020-05-1113-98/+98
| | | | | | | Move cases from teardown class to teardown in dht Change-Id: Id0cf120c6229715521ae19fd4bb00cad553d701f Signed-off-by: Sri Vignesh <sselvan@redhat.com>
* [Test] Add tc to check volume metadata-self-healkshithijiyer2020-05-111-0/+606
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Testcase steps: 1.Turn off the options self heal daemon 2.Create IO 3.Calculate arequal of the bricks and mount point 4.Bring down "brick1" process 5.Change the permissions of the directories and files 6.Change the ownership of the directories and files 7.Change the group of the directories and files 8.Bring back the brick "brick1" process 9.Execute "find . | xargs stat" from the mount point to trigger heal 10.Verify the changes in permissions are not self healed on brick1 11.Verify the changes in permissions on all bricks but brick1 12.Verify the changes in ownership are not self healed on brick1 13.Verify the changes in ownership on all the bricks but brick1 14.Verify the changes in group are not successfully self-healed on brick1 15.Verify the changes in group on all the bricks but brick1 16.Turn on the option metadata-self-heal 17.Execute "find . | xargs md5sum" from the mount point to trgger heal 18.Wait for heal to complete 19.Verify the changes in permissions are self-healed on brick1 20.Verify the changes in ownership are successfully self-healed on brick1 21.Verify the changes in group are successfully self-healed on brick1 22.Calculate arequal check on all the bricks and mount point Change-Id: Ia7fb1b272c3c6bf85093690819b68bd83efefe14 Co-authored-by: Vitalii Koriakov <vkoriako@redhat.com> Signed-off-by: Vitalii Koriakov <vkoriako@redhat.com> Signed-off-by: kshithijiyer <kshithij.ki@gmail.com>
* [Test] Add tc to check custom file xattrskshithijiyer2020-05-071-0/+256
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Description: This test case creates files at mount point and verifies custom attributes across bricks Testcase steps: 1.Create a gluster volume and start it. 2.Create file and link files. 3.Create a custom xattr for file. 4.Verify that xattr for file is displayed on mount point and bricks 5.Modify custom xattr value and verify that xattr for file is displayed on mount point and bricks 6.Verify that custom xattr is not displayed once you remove it 7.Create a custom xattr for symbolic link. 8.Verify that xattr for symbolic link is displayed on mount point and sub-volume 9.Modify custom xattr value and verify that xattr for symbolic link is displayed on mount point and bricks 10.Verify that custom xattr is not displayed once you remove it. Change-Id: Iff7360273369c77da243f2c09df2e10a0eec27ea Co-authored-by: Kartik_Burmee <kburmee@redhat.com> Signed-off-by: Kartik_Burmee <kburmee@redhat.com> Signed-off-by: kshithijiyer <kshithij.ki@gmail.com>
* [testfix][bug] Remove disperse vol type - quotaVinayak Papnoi2020-05-041-2/+1
| | | | | | | | | | | | | The test case 'tests/functional/quota/test_limit_usage_deep_dir.py' fails erratically for disperse volume. A bug [1] had been raised for the same where it was decided to remove the disperse volume type from the 'runs_on' of that test. [1] https://bugzilla.redhat.com/show_bug.cgi?id=1672983 Change-Id: Ica8f2af449225d72d1b60c2c86b20e16b80a5a5a Signed-off-by: Vinayak Papnoi <vpapnoi@redhat.com>
* [Test] Add tc to check impact of replace brick on shdkshithijiyer2020-04-291-0/+186
| | | | | | | | | | | | | | | | | | | | | | | Description: Test Script to verify the glustershd server vol file has only entries for replicate volumes. Testcase steps: 1.Create multiple volumes and start all volumes 2.Check the glustershd processes(Only 1 glustershd should be listed) 3.Do replace brick on the replicate volume 4.Confirm that the brick is replaced 5.Check the glustershd processes(Only 1 glustershd should be listed and pid should be different) 6.glustershd server vol should be updated with new bricks Change-Id: I09245c8ff6a2b31a038749643af294aa8b81a51a Co-authored-by: Vijay Avuthu <vavuthu@redhat.com>, Vitalii Koriakov <vkoriako@redhat.com> Signed-off-by: Vijay Avuthu <vavuthu@redhat.com> Signed-off-by: Vitalii Koriakov <vkoriako@redhat.com> Signed-off-by: kshithijiyer <kshithij.ki@gmail.com>
* [Test] functional/disperse: Verify remove brick operationSri Vignesh2020-04-291-0/+324
| | | | | | | | | | This test verifies remove brick operations on disperse volume. Change-Id: If4be3ffc39a8b58e4296d58b288e3843a218c468 Co-authored-by: Sunil Kumar Acharya <sheggodu@redhat.com> Signed-off-by: Sunil Kumar Acharya <sheggodu@redhat.com> Signed-off-by: Sri Vignesh <sselvan@redhat.com>
* [Test] Heal Full after deleting the files from bricksPranav2020-04-281-0/+202
| | | | | | | | | | | | | | | | | | Test case: * Create IO * Calculate arequal from mount * kill glusterd process and glustershd process on arbiter nodes * Delete data from backend from the arbiter nodes * Start glusterd process and force start the volume to bring the processes online * Check if heal is completed * Check for split-brain * Calculate arequal checksum and compare it Change-Id: I41192134530ec42db3398ae97e4f328b77e529d1 Co-authored-by: Vitalii Koriakov <vkoriako@redhat.com> Signed-off-by: Vitalii Koriakov <vkoriako@redhat.com> Signed-off-by: Pranav <prprakas@redhat.com>
* [Test+Lib] No fresh lookups on directoryBala Konda Reddy M2020-04-281-0/+183
| | | | | | | | | | | | | | | | | | | Test Steps: 1. Create a volume and set the volume option 'diagnostics.client-log-level' to DEBUG mount the volume on one client. 2. Create a directory 3. Validate the number of lookups for the directory creation from the log file. 4. Perform a new lookup of the directory 5. No new lookups should have happened on the directory, validate from the log file. 6. Bring down one subvol of the volume and repeat step 4, 5 7. Bring down one brick from the online bricks and repeat step 4, 5 8. Start the volume with force and wait for all process to be online. Change-Id: I162766837fd7e61625238a669c4050c2ec9c8a8b Signed-off-by: Bala Konda Reddy M <bmekala@redhat.com>
* [TestFix] Add code to override the volume configsayaleeraut2020-04-231-38/+73
| | | | | | | | | | | | | | | Changing the distribute count to 4 for the volume type distributed-replicated or distributed-dispersed, as earlier with distribute count 2, after remove-brick, the dist-rep & dist-disp volumes were converted to pure rep or pure dispersed, which caused "layout not complete" error as with the DHT pass-through feature layout is not set on bricks if volume type is pure replicated/pure dispersed on gluster version 6.0 Adding distributed-arbiter volume type and have added code to override its configuration as well. Change-Id: Ic7a3404ed49d24f956de33f7bd5ca8ea61297e5b Signed-off-by: sayaleeraut <saraut@redhat.com>
* [Test] Add tc to check get-state when brick is killedPranav2020-04-231-0/+124
| | | | | | | | | Test case verifies whether the gluster get-state shows the proper brick status in the output. The test case checks the brick status when the brick is up and also after killing the brick process. It also verifies whether the other bricks are up when a particular brick process is killed. Change-Id: I9801249d25be2817104194bb0a8f6a16271d662a Signed-off-by: Pranav <prprakas@redhat.com>