Lustre 2.10.4 Changelog

Server support for kernels:

• RHEL 7.5 (3.10.0-862.2.3.el7)

Client support for unpatched kernels:

• RHEL 6.9 (2.6.32-696.28.1.el6)
• RHEL 7.5 (3.10.0-862.2.3.el7)
• SLES12 SP3 (4.4.126-94.22)
• Ubuntu 16.04 (4.4.0-116)

Recommended e2fsprogs version:

• v1.42.13.wc6

Interoperability Support:

• Clients & Servers: Latest 2.9.X

Issues fixed between 2.10.3 and 2.10.4

New Feature

• LU-10030: add native lfs tools support for project quota

Improvement

• LU-6051: "lfs_migrate" improvements
• LU-6353: Push Wireshark Support Upstream
• LU-7251: reduce commit callbacks in OSP
• LU-8649: Print console message in recovery when waiting for first client
• LU-9796: Speedup file creation under heavy concurrency
• LU-9946: Have fileset on default nodemap
• LU-9950: add support for Ubuntu(debian) arm64
• LU-10516: Update recommended e2fsprogs version
• LU-10676: lustre-zfs-dkms rpm should Provide lustre-dkms for easy upgrades
• LU-10864: update changelog kernel version for ubuntu1604

Bug

• LU-5024: (mdc_lib.c:163:mdc_pack_name()) ASSERTION( cpy_len == name_len && lu_name_is_valid_2(buf, cpy_len) ) failed:
• LU-5152: Can't enforce block quota when unprivileged user change group
• LU-5163: (lu_object.h:852:lu_object_attr()) ASSERTION( ((o)->lo_header->loh_attr & LOHA_EXISTS) != 0 ) failed
• LU-5490: sanity test_133d: FAIL: samedir_rename_size error
• LU-5680: sanity-hsm uses /etc/motd (which is probably of size 0)
• LU-7001: osp_sync.c: 1139: osp_sync_thread
• LU-7746: skip test of new functionality on upstream client
• LU-7988: HSM: high lock contention for cdt_llog_lock
• LU-8264: lfs setstripe without -p pool_name doesn't inherit pool from parent/ROOT directory
• LU-8346: conf-sanity test_93: test failed to respond and timed out
• LU-8653: broken inheritance of default striping
• LU-8856: ZFS-MDT 100% full. Cannot delete files.
• LU-8912: nodemap: error: nodemap_add_range: nodemap ranges must be contiguous
• LU-8913: nodemap: wildcard '*@<net>' is broken
• LU-8990: Failback LBUG lod_device_free()) ASSERTION( atomic_read(&lu->ld_ref)
• LU-9052: lod_verify_md_striping()) lsh-MDT0000-mdtlov: invalid lmv_user_md: magic = cd20cd0
• LU-9135: sanity test_313: osp_sync.c:571:osp_sync_interpret()) LBUG
• LU-9145: When Shared Key feature is active, Nodemap admin property allows more access
• LU-9158: sanity-quota, test_12b: rebalancing failed
• LU-9372: OOM happens on OSS during Lustre recovery for more than 5000 clients
• LU-9411: sanity test_27D: interop 2.9.0<->master llapi_layout_test 30, 31 failure
• LU-9437: sanity-lfsck test_33: only 0 of 4 MDTs are in completed
• LU-9444: conf-sanity test_87: Create file with 3 components failed
• LU-9551: I/O errors when lustre uses multipath devices
• LU-9664: sanity-hsm test_9a: @@@@@@ FAIL: mds4 CDT state is not 'disabled'
• LU-9735: Sles12Sp2 and 2.9 getcwd() sometimes fails
• LU-9833: New gcc compiler generates a lot of format-truncation warnings
• LU-9836: Issues with 2.10 upgrade and files missing LMAC_FID_ON_OST flag
• LU-9844: building lustre utils lsnapshot.c fails for arch ppc64le
• LU-9874: simplify project transfer codes
• LU-9887: sanity-lfsck test_9a: FAIL: (4) Got speed 952, expected less than 144
• LU-9892: parallel-scale-nfsv3 no sub tests failed: setup nfs failed!
• LU-9967: sanity test_60a: Module llog_test is in use
• LU-9972: Performance regressions on unique directory removal
• LU-9983: LBUG llog_osd.c:327:llog_osd_declare_write_rec() - all DNE MDS
• LU-9998: Default partition setup is not optimal for best metadata performance
• LU-10005: File creation to slave MDT is much slower than primary MDT on DNE1 configuration
• LU-10011: suppress annoying messages for lfs quota
• LU-10045: sanity-lfsck no sub tests failed
• LU-10052: replay-single test_20b fails with 'after 4096 > before 3072'
• LU-10161: lbuild: Failed to determine the default route interface
• LU-10199: sanity test_56xb: hard links individually migrated (12 != 11)
• LU-10212: operation ost_read to node [*] failed: rc = -116
• LU-10214: Example `lnet_routes.conf` syntax for gateway range is not supported
• LU-10224: recovery-small test_57: timeout
• LU-10226: sanity-sec test_27: fileset not cleared on default nodemap
• LU-10237: "ls" hangs on a particular directory
• LU-10244: LustreError: 201288:0:(osc_page.c:336:osc_page_delete()) ASSERTION( (!(value != ((void *)0)) || (value == opg)) ) failed:
• LU-10290: sanity-sec test_27: fileset info still not cleared on nodemap
• LU-10297: parallel-scale-nfsv4 test_metabench: ASSERTION( nfound <= inuse->op_count ) failed
• LU-10316: Interop 2.7.x <->2.10.2 sanity test_77c: FAIL: dump content does not match on OSS
• LU-10321: MDS - umount hangs during failback
• LU-10327: sanity-hsm, test_24c: cannot chown '/mnt/lustre/d24c.sanity-hsm/f24c.sanity-hsm' to '500:nobody'
• LU-10331: mds-survey test_1: mds-survey failed
• LU-10341: kkuc shoud send HALs only to suitable CTs
• LU-10350: ost-pools test 1n fails with 'failed to write to /mnt/lustre/d1n.ost-pools/file: 1'
• LU-10356: CLIO simplification broke a direct IO sometimes
• LU-10357: ll_ioc_copy_{start,end}() depend on search_inode_for_lustre() which is bad
• LU-10364: add version check to sanity test_255b
• LU-10368: disk quota OST rebalancing issues
• LU-10392: LustreError: 82980:0:(fid_handler.c:329:__seq_server_alloc_meta()) srv-lglossy-MDT0002: Allocated super-sequence failed: rc = -115
• LU-10421: mds-survey test 1: Timeout occurred after 426 mins, last suite running was mds-survey, restarting cluster to continue tests
• LU-10422: layout LFSCK try to fix consistent owner unexpectedly
• LU-10437: sanity-pfl test_8: dbench failed
• LU-10444: l_getidentity keeps remount /sys/kernel/debug and reverting permissions.
• LU-10449: TBF policy can't be set correctly
• LU-10462: lfs setstripe --component-add fails to recognize -c -1 as all targets
• LU-10463: Poor write performance periodically on repeated test runs
• LU-10505: sanity test_56w: '/usr/bin/lfs getstripe -c /mnt/lustre/d56w.sanityw/file1 wrong: found 7, expected 6
• LU-10531: GSS, Shared Key and Kerberos support broken in master and lustre 2.10
• LU-10541: parallel-scale-nfsv3 test racer_on_nfs hangs with BUG: unable to handle kernel NULL pointer dereference
• LU-10560: Fixes for 4.14 kernel
• LU-10563: kernel update [RHEL7.4 3.10.0-693.17.1.el7]
• LU-10564: kernel update [RHEL6.9 2.6.32-696.20.1.el6]
• LU-10565: Handle some small features in upstream
• LU-10566: parallel-scale-nfsv4 test_metabench: mkdir: cannot create directory on Read-only file system
• LU-10575: No mdsrate on Ubuntu client
• LU-10598: Ignore IGIF formatted last_id
• LU-10603: ptlrpc_lprocfs_req_buffers_max_fops unused
• LU-10634: kernel update [SLES12 SP3 4.4.114-94.11]
• LU-10650: cslco1705 crash: dt_statfs()) ASSERTION( dev ) failed: LBUG, Pid: 3372, comm: lctl
• LU-10653: kernel update [SLES12 SP2 4.4.114-92.64]
• LU-10655: llmout.sh and llmountcleanup.sh generates meaningless output: ssh exited with exit code 1
• LU-10656: the reference of obd_export was leaked in ptlrpc_connect_interpret
• LU-10663: obdfilter-survey
• LU-10682: o2iblnd txs can be dropped silently
• LU-10703: All mds Nodemap filset will be cleared when do some nodemap operations (lustre 2.10.3)
• LU-10707: TCP eth routed LNet traffic broken
• LU-10731: kernel update [SLES12 SP3 4.4.114-94.14]
• LU-10735: kernel update [SLES12 SP2 4.4.114-92.67]
• LU-10759: sanity test 27F fails with 'wrong stripecount' when run with a single OST
• LU-10761: NOT create REMOTE_PARENT_DIR on OST
• LU-10764: Incorrect code in ct_archive
• LU-10769: osd-zfs deadlock on osd_object::oo_guard
• LU-10773: soft lockup when remove changelog
• LU-10783: kernel update [RHEL7.4 3.10.0-693.21.1.el7]
• LU-10794: lfs quota did not work well for grace time output
• LU-10795: wrong skipping of reintegration
• LU-10802: mismatch problem for wildcard in jobid TBF
• LU-10803: req_buffers_max and req_history_max setting problems
• LU-10813: kernel update [RHEL6.9 2.6.32-696.23.1.el6]
• LU-10837: no bitmap check if block bitmap is uninitialized
• LU-10845: kernel update [SLES12 SP3 4.4.120-94.17]
• LU-10847: kernel update [SLES12 SP2 4.4.120-92.70]
• LU-10859: Deadlock with heavy memory pressure
• LU-10887: 2 MDTs stuck in WAITING
• LU-10897: kernel upgrade [RHEL7.5 3.10.0-862.2.3.el7]
• LU-10912: avoid triggering LU-9856 for older servers
• LU-10944: kernel update [SLES12 SP3 4.4.126-94.22]
• LU-11008: kernel update [RHEL6.9 2.6.32-696.28.1.el6]
• LU-11024: Broken inode accounting of MDT on ZFS
• LU-11026: lustre-dkms should require patch or quilt

Task

• LU-9761: Add ldiskfs support to dkms for patchless kernel
• LU-10377: Update ZFS Version to 0.7.6
• LU-10460: Evaluate two LLNL patches for upstream
• LU-10569: Include proper Lustre header files in lustre-dev_*.deb
• LU-10901: Update ZFS Version to 0.7.8
• LU-11019: Update ZFS Version to 0.7.9

Technical task

• LU-10248: Need to update PFID of OST objects after layout change

Question/Request

• LU-9594: sanity test cases 42a-d 45 68 76 407 65ic remove from ALWAYS_EXCEPT