Ceph » CephFS

Link to the job https://pulpito.ceph.com/xiubli-2024-08-07_00:51:19-fs-wip-xiubli-testing-20240806.124812-reef-distro-default-smithi/7841220

2024-08-08T11:42:44.725 INFO:tasks.cephfs_test_runner:======================================================================
2024-08-08T11:42:44.725 INFO:tasks.cephfs_test_runner:FAIL: test_fscrypt_journal_recovery (tasks.cephfs.test_fscrypt.TestFSCryptRecovery)
2024-08-08T11:42:44.725 INFO:tasks.cephfs_test_runner:That alternate_name can be recovered from the journal.
2024-08-08T11:42:44.725 INFO:tasks.cephfs_test_runner:----------------------------------------------------------------------
2024-08-08T11:42:44.726 INFO:tasks.cephfs_test_runner:Traceback (most recent call last):
2024-08-08T11:42:44.726 INFO:tasks.cephfs_test_runner:  File "/home/teuthworker/src/github.com_ceph_ceph-c_5536d05be0501ca23bda2e7c129ea5d1781edf43/qa/tasks/cephfs/test_fscrypt.py", line 86, in test_fscrypt_journal_recovery
2024-08-08T11:42:44.726 INFO:tasks.cephfs_test_runner:    verify_alternate_name()
2024-08-08T11:42:44.726 INFO:tasks.cephfs_test_runner:  File "/home/teuthworker/src/github.com_ceph_ceph-c_5536d05be0501ca23bda2e7c129ea5d1781edf43/qa/tasks/cephfs/test_fscrypt.py", line 64, in verify_alternate_name
2024-08-08T11:42:44.726 INFO:tasks.cephfs_test_runner:    self.assertEqual(len(dirfrags), 1)
2024-08-08T11:42:44.726 INFO:tasks.cephfs_test_runner:AssertionError: 0 != 1

From test_fscrypt's test_fscrypt_journal_recovery -

        def verify_alternate_name():
            J = self.fs.read_cache("/dir", depth=0)
            self.assertEqual(len(J), 1)
            inode = J[0]
            dirfrags = inode['dirfrags']
            self.assertEqual(len(dirfrags), 1)

From output of dump tree which has no dirfrags -

2024-08-08T11:42:34.643 DEBUG:teuthology.orchestra.run.smithi029:> sudo adjust-ulimits ceph-coverage /home/ubuntu/cephtest/archive/coverage timeout 300 ceph --cluster ceph --admin-daemon /var/run/ceph/ceph-mds.a.asok --format=json dump tree /dir 0
...
    "dirfrags": [],
    "dirfragtree": {
      "splits": []
    },

From qa/tasks/cephfs/filesystem.py -

    def read_cache(self, path, depth=None, rank=None):
        cmd = ["dump", "tree", path]

Another issue occurred during FSCryptTestCase.tearDown() -

2024-08-08T11:42:34.759 DEBUG:teuthology.orchestra.run.smithi165:> (cd /home/ubuntu/cephtest/mnt.0 && exec bash -c 'sudo fscrypt purge --force --quiet /home/ubuntu/cephtest/mnt.0')
2024-08-08T11:42:35.096 INFO:tasks.ceph.mds.a.smithi029.stderr:2024-08-08T11:42:35.096+0000 7fd6a80ab640 -1 log_channel(cluster) log [ERR] : bad backtrace on directory inode 0x10000000001
2024-08-08T11:42:35.099 INFO:tasks.ceph.mds.a.smithi029.stderr:2024-08-08T11:42:35.098+0000 7fd6a80ab640 -1 log_channel(cluster) log [ERR] : bad backtrace on directory inode 0x10000000002
...
2024-08-08T11:42:44.726 INFO:tasks.cephfs_test_runner:Traceback (most recent call last):
...
2024-08-08T11:43:13.640 INFO:teuthology.orchestra.run.smithi029.stdout:2024-08-08T11:42:35.096384+0000 mds.a (mds.0) 1 : cluster [ERR] bad backtrace on directory inode 0x10000000001

From test_fscrypt -

    def tearDown(self):
        self.mount_a.run_shell_payload(f"sudo fscrypt purge --force --quiet {self.mount_a.hostfs_mntpt}")

        super().tearDown()