Skip to content

v0.7.0-rc4

Pre-release
Pre-release
Compare
Choose a tag to compare
@behlendorf behlendorf released this 05 May 19:15
· 6793 commits to master since this release
zfs-0.7.0-rc4
This tag was signed with the committer’s verified signature.
behlendorf Brian Behlendorf
GPG key ID: 0AB9E991C6AF658B
Learn about vigilant mode.
a0e8401

Supported Kernels

  • Compatible with 2.6.32 - 4.11 Linux kernels.

New Features

  • Added script support to zpool iostat -c extended output
  • Added feature@multi_vdev_crash_dump compatibility
  • Added GZIP compression offloading with QAT accelerator

Performance

  • Fixed performance of 1MB writes on certain RAID-Z configurations (OpenZFS 8005)
  • Fixed lseek(SEEK_HOLE) performance for dirty files
  • Fixed large file delete can starving out write ops (OpenZFS 6569)
  • Improved ARC memory reclaim behavior
  • Improved asynchronous ZVOL performance
  • Allow higher ashift values (up to 16)
  • Multi-threaded spa_sync() (OpenZFS 7968)
  • Increased zfs_vdev_async_write_min_active (to 2) which can speed up rebuilds
  • Increased indirect block size (OpenZFS 7104)

Bug Fixes

  • Fixed immediately update label 2,3 uberblocks when vdev expands
  • Fixed human-readable sizes in command line utilities
  • Fixed page size alignment for large buffers
  • Fixed zfs_dirty_data_max_max limit (4G)
  • Fixed incorrect compressed send streams
  • Fixed zdb -e regression for active cacheless pools
  • Fixed vdev_id failure due to multipath -l bug
  • Fixed leak in send_iterate_fs()
  • Fixed large block/gzip/raidz boot pools allowed
  • Fixed ZVOL BLKFLSBUF ioctl invalidates cache
  • Fixed block device cache invalidation during zpool import|labelclear
  • Fixed header inclusions for standards conformance
  • Fixed size inflation in spa_get_worst_case_asize
  • Fixed panic destroying a metaslab deferred range tree (OpenZFS 8023)
  • Fixed zpool create accepts raidz and mirrors with similar redundancy
  • Fixed zpool list reporting 16.0e for expandsz (OpenZFS 7885)
  • Fixed check ashift validity in zpool add
  • Fixed wrong offset args in vdev_cache_write()
  • Fixed 'zdb -o' segmentation fault
  • Fixed zfs_znode_alloc() failure in zfs_mknode() with retry
  • Fixed race in rollback, zil close, and zil flush (OpenZFS 3821)
  • Fixed zpool status -v error message
  • Fixed NFS snapdir automounting
  • Fixed harmless "BARRIER is deprecated" kernel warning on Centos 6.8
  • Fixed ARC space accounting leak (OpenZFS 7867)
  • Fixed loop device becoming read-only for Linux kernel older than 4.1
  • Fixed dsl_dataset_rollback_sync() may try to free already free blocks (OpenZFS 7199)
  • Fixed off by one in zpl_lookup
  • Fixed enclosure LED handling
  • Fixed zfs receive of deduplicated streams
  • Fixed zfs get behavior with bookmarks (OpenZFS 7386)
  • Added zdb(8) lookup files by path (OpenZFS 6410)
  • Added zdb(8) only dump unique configurations and uberblocks
  • Added zdb -G to print zfs_dbgmsg's
  • Added zdb -ll to print the nvlist payload stats
  • Added constrained paths to the ZFS Test Suite (OpenZFS 7290)
  • Added test framework for the ZED to the ZFS Test Suite (auto-online, auto-replace)
  • Added new test cases and enabled additional existing test cases in the ZFS Test Suite
  • Added debug and console logs dumping to the ZFS Test Suite
  • Allow c99 when building ZFS in the kernel tree
  • Applied 106 upstream OpenZFS patches
  • Assorted documentation and man page updates

This tag is provided for evaluation and testing of the upcoming 0.7.0 release.

Assets 5
Loading