-
Tue Nov 20 2018 Tiger Yang <tiger.yang@oracle.com> - 3.12.15-3.el7
- fix CVE bugs, upgarde to 3.12.15-3 (Tiger Yang)
- lock: Do not allow meta-lock count to be more than one (Susant Palai) {CVE-2018-14660}
- index: prevent arbitrary file creation outside entry-changes folder (Ravishankar N) {CVE-2018-14654}
- all: fix the format string exceptions (Amar Tumballi) {CVE-2018-14661}
- io-stats: prevent taking file dump on server side (Amar Tumballi) {CVE-2018-14659}
- protocol: remove the option 'verify-volfile-checksum' (Amar Tumballi) {CVE-2018-14653}
- glusterd-handshake: prevent a buffer overflow (Amar Tumballi) {CVE-2018-14653}
- server: don't allow '/' in basename (Amar Tumballi) {CVE-2018-14651}
- features/locks: add buffer overflow checks in pl_getxattr (Ravishankar N) {CVE-2018-14652}
-
Mon Oct 22 2018 Tiger Yang <tiger.yang@oracle.com> - 3.12.15-2.el7
- set vendor string to "Oracle America" (Tiger Yang)
-
Mon Oct 15 2018 Tiger Yang <tiger.yang@oracle.com> - 3.12.15-1.el7
- upgrade to 3.12.15
-
Mon Oct 15 2018 Tiger Yang <tiger.yang@oracle.com> - 3.12.15-1
- samba: Gluster hooks for Samba config without kernel share mode
- Merge tag 'v3.12.15' into oracle-gluster-3.12
- Release notes for Gluster 3.12.15
- afr: prevent winding inodelks twice for arbiter volumes
- cluster/afr: Delegate metadata heal with pending xattrs to SHD
- cluster/afr: Delegate name-heal when possible
- dht: Fill first_up_subvol before use in dht_opendir
- packaging: manual systemctl daemon reload required after install
- afr: fix incorrect reporting of directory split-brain
- glusterd: volume inode/fd status broken with brick mux
- doc: Release notes for v3.12.4
- rpc: handle EAGAIN when SSL_ERROR_SYSCALL is returned
- protocol: don't use alloca
- posix: disable open/read/write on special files
- io-stats: dump io-stats info in /var/run/gluster
- server-protocol: don't allow '../' path in 'name'
- dict: handle negative key/value length while unserialize
- posix: remove not supported get/set content
- Release notes for Gluster 3.12.13
- Bash integration script should namespace variables
- gfapi : Handle the path == "" glfs_resolve_at
- cluster/afr: Fix dict-leak in pre-op
- glusterd: _is_prefix should handle 0-length paths
- posix: check before removing stale symlink
- Release notes for 3.12.12
- afr: don't update readables if inode refresh failed on all child
- afr: heal gfids when file is not present on all bricks
- afr: fix bug-1363721.t failure
- afr: add quorum checks in pre-op
- afr: capture the correct errno in post-op quorum check
- cluster/dht: act as passthrough for renames on single child DHT
- glusterfsd: Do not process GLUSTERD_BRICK_XLATOR_OP if graph is
- afr: add quorum checks in post-op
- glusterd: gluster v status is showing wrong status for glustersh
- afr: don't treat all cases all bricks being blamed as split-brai
- storage/posix: Fix posix_symlinks_match()
- cluster/dht: Remove EIO from dht_inode_missing
- doc: Updated release notes for 3.12.11
- glusterfs: access trusted peer group via remote-host command
- storage/posix: Handle ENOSPC correctly in zero_fill
- gcron: create the lockfile if it is missing
-
Fri Jun 22 2018 Tiger Yang <tiger.yang@oracle.com> - 3.12.10-1
- glusterfs: access trusted peer group via remote-host command
- Release notes for 3.12.10
- protocol/server: Fix xdata leak in seek fop
- geo-rep/scheduler: Fix crash
- glusterd/geo-rep: Fix glusterd crash
- geo-rep: Fix upgrade issue
- geo-rep: Fix syncing of symlink
- use awk to get a specific line from the output instead of cut
- gfapi: various broken symbol versions
- core: FreeBSD has pthread_set_name_np() (versus pthread_setname_
- test: Marking test case bug-1309462.t as bad
-
Tue May 15 2018 Tiger Yang <tiger.yang@oracle.com> - 3.12.9-1
- doc: Added release notes for 3.12.9 release
- libglusterfs: fix comparison of a NULL dict with a non-NULL dict
- server/auth: add option for strict authentication
- shared storage: Prevent mounting shared storage from non-trusted client
- cluster/dht: Handle file migrations when brick down
- cluster/dht: Wind open to all subvols
- cluster/afr: Fixing the flaws in arbiter becoming source patch
- cluster/afr: Fix for arbiter becoming source
- Release notes for 3.12.8
- features/index: Choose different base file on EMLINK error
- timer: Fix possible race during cleanup
- cluster/dht: Skipped files are not treated as errors
- cluster/afr: Prevent ping-event handling on shd
- cluster/ec: send list-node-uuids request to all subvolumes
- cluster/ec: Change default read policy to gfid-hash
- cluster/ec: avoid delays in self-heal
- extras/hooks: Fix S10selinux-label-brick.sh hook script
- glusterfsd: Memleak in glusterfsd process while brick mux is on
- glusterd: import volumes in separate synctask
- cluster/dht: ENOSPC will not fail rebalance
- Release notes for 3.12.7
- glusterd: optimize glusterd import volumes code path
- cluster/afr: Fail open on split-brain
- glusterd/store: handle the case of fsid being set to 0
- cluster/dht: Handle single dht child in dht_lookup
- cluster/dht: Ignore ENODATA from getxattr for posix acls
- cluster/dht: Fixed a typo
- protocol/server: Backport patch to reduce duplicate code in server-rpc-fops.c
-
Tue Feb 27 2018 Tiger Yang <tiger.yang@oracle.com> - 3.12.6-1
- doc : release-notes for GlusterFS-3.12.6
- glusterd: fix tier-enabled flag op-version check
- tests/dht: Non-root can delete stale linkto files
- nfs: Fixing the failure in bug-974972.t test case
- cluster/dht: Skip '..' for the volume root dir
- rpc: Showing some unusual timer error logs during brick stop
- cluster/dht: Cleanup on fallocate failure
- cluster/dht: Unlink linkto files as root
- cluster/dht: Fixed leak in dht_populate_inode_for_dentry
- tests: mark test bad
- tests: Spurious failure multiplex-limit-issue-151.t
- cluster/afr: remove unnecessary child_up initialization
- cluster/dht: Add migration checks to dht_(f)xattrop
- tests: Fix a broken test case
- cluster/ec: OpenFD heal implementation for EC
- glusterd: process pmap sign in only when port is marked as free
- selinux-xlator : validate dict before calling dict_rename_key()
- geo-rep: Fix data sync issue during hardlink, rename
-
Tue Jan 16 2018 Tiger Yang <tiger.yang@oracle.com> - 3.12.5-1
- doc : release-notes for GlusterFS-3.12.5
- glusterd: connect to an existing brick process when qourum status is NOT_APPLICABLE_QUORUM
- cluster/dht: Use percentages for space check
- quota: fixes issue in quota.conf when setting large number of limits
- Revert "mount/fuse: report ESTALE as ENOENT"
- cli: Fixed a use_after_free
- glusterd: Nullify pmap entry for bricks belonging to same port
- mount/fuse: never fail open(dir) with ENOENT
- performance/write-behind: fix bug while handling short writes
- rpc-transport/rdma: Add a mutex for the list of RDMA Memory Region(MR) access
- eventsapi: JWT signing without external dependency
- mount/fuse: use fstat in getattr implementation if any opened fd is available
-
Mon Dec 25 2017 Tiger Yang <tiger.yang@oracle.com> - 3.12.4-1
- doc : release-notes for GlusterFS-3.12.4
- glusterd: Free up svc->conn on volume delete
- core/memacct: save allocs in mem_acct_rec list
- Disable gfid2path by default on NetBSD
- cluster/dht: don't overfill the buffer in readdir(p)
- cluster/dht: populate inode in dentry for single subvolume dht
- cli: remove summary option from heal info command
- tests: fix bug-1483058-replace-brick-quorum-validation.t spurious failure
- glusterd: display gluster volume status, when quorum type is server
- cluster/afr: Print heal info split-brain output in stream fashion
- cluster/afr: Honor default timeout of 5min for analyzing split-brain files
- features/locks: Fix memory leaks
- cluster/dht: make rebalance use truncate incase
- cluster/dht: Don't set ACLs on linkto file
-
Mon Dec 11 2017 Tiger Yang <tiger.yang@oracle.com> - 3.12.3-1
- (tag: v3.12.3)doc : release-notes for GlusterFS-3.12.3
- glusterd: restart the brick if qorum status is NOT_APPLICABLE_QUORUM
- cli: correct rebalance status elapsed check
- eventsapi: HTTPS support for Webhooks
- md-cache: avoid checking the xattr value buffer with string functions.
- cli: gluster help changes
- glusterd : introduce timer in mgmt_v3_lock
- protocol/server: fix the comparision logic in case of subdir mount
- protocol/client: handle the subdir handshake properly for add-brick
- md-cache: Use correct xattr keynames for virtual glusterfs ACLs.
- glusterd: clean up portmap on brick disconnect
- glusterd: fix brick restart parallelism
- glusterd: delete source brick only once in reset-brick commit force
- glusterd: persist brickinfo's port change into glusterd's store
- features/worm: Adding check to newloc when doing rename
- glusterd: documenting server.allow-insecure
- cli: volume status command is missing in man page
- glusterfsd: Dereferencing the null pointer
- build: make it possible to build cleanly 2x in a row
- eventsapi: Add JWT signing support
- libgfchangelog: Fix possible null pointer dereference
- cluster/dht: fix crash when deleting directories
- protocol-auth: use the proper validation method
- gfapi: set lkowner in glfd
- geo-rep: Filter out volume-mark xattr
- mount/fuse : Fix parsing of vol_id for snapshot volume
- (tag: v3.12.2)doc : release-notes for GlusterFS-3.12.2
- cluster/ec: Improve performance with xattrop update
- glusterd:Marking all the brick status as stopped when a process goes down in brick multiplexing
- geo-rep: Fix status transition
- cluster/afr: Make choose-local "reconfigurable"
- geo-rep/scheduler: Add validation for session existence
- fuse/readdirp: Remove need_lookup from fuse_readdirp_cbk
- geo-rep: Add EINTR to retry list while doing readlink
- geo-rep: Add ENODATA to retry list on gfid getxattr
- geo-rep: Fix passive brick's last sync time
- geo-rep: Add ENOTSUP error to retry list
- cluster/dht: Don't store the entire uuid for subvols
- glusterd: disallow replace brick for dist only volumes
- cli/afr: gluster volume heal info "healed" command output is not appropriate
- glusterd: fix invalid memory reference returned
- afr: heal gfid as a part of entry heal
- glusterd: fix client io-threads option for replicate volumes
- glusterd: spelling errors reported by Debian maintainer
- Posix: fix additional inode ref
- rpc: TLSv1_2_method() is deprecated in OpenSSL-1.1
- cluster/afr: Sending subvol up/down events when subvol comes up or goes down
- cli: Bypass the confirmation question creating replica 2 volume with force
- geo-rep/scheduler: Fix '--no-color' help msg
- features/locks: Maintain separation of lock->client_pid, flock->l_pid
- afr: don't check for file size in afr_mark_source_sinks_if_file_empty
- glusterd: retrieve uuid under mutex lock
- afr: auto-resolve split-brains for zero-byte files
- mount/fuse: Make event-history feature configurable
- eventsapi: Fix issue with CLIENT_CONNECT event
- events : Add brick list in EVENT_VOLUME_CREATE
- features/shard: Change default shard-block-size to 64MB
- mount/fuse: Include sub-directory in source argument for mount()
- (tag: v3.12.1)doc : release-notes for GlusterFS-3.12.1
- afr: heal metadata in discover code path
- event/epoll: don't call handler for events received after a pollerr
- arbiter: return ENOSYS for 'non readable' FOPs
- dht: add FOP check to dht_file_setattr_cbk
- cluster/dht: Log files skipped by rebalance
- cluster/dht: Aggregate xattrs only for dirs in dht_discover_cbk
- perf/qr: Use a ref-ed data to extract content
- rpc: destroy transport after client_t
- afr: check validity of afr_reply
- features/shard: Return aggregated size in stbuf of LINK fop
- features/shard: Increment counts in locks
- scripts: mount.glusterfs contains non-portable bashisms
- (tag: v3.12.0)feature/posix: Enabled gfid2path by default
- doc: Finalized release notes for 3.12.0 release