From 545a7ce6762a1b3a7b989b43a9d18b5b1b299df0 Mon Sep 17 00:00:00 2001 From: Susant Palai Date: Thu, 18 Jan 2018 13:06:12 +0530 Subject: cluster/dht: avoid overwriting client writes during migration For more details on this issue see https://github.com/gluster/glusterfs/issues/308 Solution: This is a restrictive solution where a file will not be migrated if a client writes to it during the migration. This does not check if the writes from the rebalance and the client actually do overlap. If dht_writev_cbk finds that the file is being migrated (PHASE1) it will set an xattr on the destination file indicating the file was updated by a non-rebalance client. Rebalance checks if any other client has written to the dst file and aborts the file migration if it finds the xattr. updates gluster/glusterfs#308 Change-Id: I73aec28bc9dbb8da57c7425ec88c6b6af0fbc9dd Signed-off-by: Susant Palai Signed-off-by: Raghavendra G Signed-off-by: N Balachandran --- xlators/storage/posix/src/posix-entry-ops.c | 12 ++++++++ xlators/storage/posix/src/posix-handle.h | 4 +++ xlators/storage/posix/src/posix-helpers.c | 42 ++++++++++++++++++++++++++ xlators/storage/posix/src/posix-inode-fd-ops.c | 9 ++++++ 4 files changed, 67 insertions(+) (limited to 'xlators/storage') diff --git a/xlators/storage/posix/src/posix-entry-ops.c b/xlators/storage/posix/src/posix-entry-ops.c index 4d7ed5be7c..41d8c873b4 100644 --- a/xlators/storage/posix/src/posix-entry-ops.c +++ b/xlators/storage/posix/src/posix-entry-ops.c @@ -180,6 +180,7 @@ posix_lookup (call_frame_t *frame, xlator_t *this, int32_t nlink_samepgfid = 0; struct posix_private *priv = NULL; posix_inode_ctx_t *ctx = NULL; + int ret = 0; VALIDATE_OR_GOTO (frame, out); VALIDATE_OR_GOTO (this, out); @@ -248,6 +249,17 @@ posix_lookup (call_frame_t *frame, xlator_t *this, if (xdata && (op_ret == 0)) { xattr = posix_xattr_fill (this, real_path, loc, NULL, -1, xdata, &buf); + + if (dict_get (xdata, GF_CLEAN_WRITE_PROTECTION)) { + ret = sys_lremovexattr (real_path, + GF_PROTECT_FROM_EXTERNAL_WRITES); + if (ret == -1 && (errno != ENODATA && errno != ENOATTR)) + gf_msg (this->name, GF_LOG_ERROR, + P_MSG_XATTR_NOT_REMOVED, errno, + "removexattr failed. key %s path %s", + GF_PROTECT_FROM_EXTERNAL_WRITES, + loc->path); + } } if (priv->update_pgfid_nlinks) { diff --git a/xlators/storage/posix/src/posix-handle.h b/xlators/storage/posix/src/posix-handle.h index 97186f91e6..8a07cf2b57 100644 --- a/xlators/storage/posix/src/posix-handle.h +++ b/xlators/storage/posix/src/posix-handle.h @@ -186,4 +186,8 @@ int posix_create_link_if_gfid_exists (xlator_t *this, uuid_t gfid, char *real_path, inode_table_t *itable); +int +posix_check_internal_writes (xlator_t *this, fd_t *fd, int sysfd, + dict_t *xdata); + #endif /* !_POSIX_HANDLE_H */ diff --git a/xlators/storage/posix/src/posix-helpers.c b/xlators/storage/posix/src/posix-helpers.c index 90afced604..56c9b4afe9 100644 --- a/xlators/storage/posix/src/posix-helpers.c +++ b/xlators/storage/posix/src/posix-helpers.c @@ -2739,3 +2739,45 @@ posix_override_umask (mode_t mode, mode_t mode_bit) gf_msg_debug ("posix", 0, "The value of mode is %u", mode); return mode; } + +int +posix_check_internal_writes (xlator_t *this, fd_t *fd, int sysfd, + dict_t *xdata) +{ + int ret = 0; + size_t xattrsize = 0; + data_t *val = NULL; + + LOCK (&fd->inode->lock); + { + val = dict_get (xdata, GF_PROTECT_FROM_EXTERNAL_WRITES); + if (val) { + ret = sys_fsetxattr (sysfd, + GF_PROTECT_FROM_EXTERNAL_WRITES, + val->data, val->len, 0); + if (ret == -1) { + gf_msg (this->name, GF_LOG_ERROR, + P_MSG_XATTR_FAILED, + errno, "setxattr failed key %s", + GF_PROTECT_FROM_EXTERNAL_WRITES); + } + + goto out; + } + + if (dict_get (xdata, GF_AVOID_OVERWRITE)) { + xattrsize = sys_fgetxattr (sysfd, + GF_PROTECT_FROM_EXTERNAL_WRITES, + NULL, 0); + if ((xattrsize == -1) && ((errno == ENOATTR) || + (errno == ENODATA))) { + ret = 0; + } else { + ret = -1; + } + } + } +out: + UNLOCK (&fd->inode->lock); + return ret; +} diff --git a/xlators/storage/posix/src/posix-inode-fd-ops.c b/xlators/storage/posix/src/posix-inode-fd-ops.c index c90bc6c438..812cf79287 100644 --- a/xlators/storage/posix/src/posix-inode-fd-ops.c +++ b/xlators/storage/posix/src/posix-inode-fd-ops.c @@ -1579,6 +1579,15 @@ posix_writev (call_frame_t *frame, xlator_t *this, fd_t *fd, _fd = pfd->fd; + ret = posix_check_internal_writes (this, fd, _fd, xdata); + if (ret < 0) { + gf_msg (this->name, GF_LOG_ERROR, 0, 0, + "possible overwrite from internal client, fd=%p", fd); + op_ret = -1; + op_errno = EBUSY; + goto out; + } + if (xdata) { if (dict_get (xdata, GLUSTERFS_WRITE_IS_APPEND)) write_append = _gf_true; -- cgit