diff options
author | Pranith Kumar K <pkarampu@redhat.com> | 2013-12-15 08:05:04 +0530 |
---|---|---|
committer | Pranith Kumar Karampuri <pkarampu@redhat.com> | 2014-10-15 01:22:02 -0700 |
commit | 7e8eefca2caaaa61e1b31b747384f660c595d9c9 (patch) | |
tree | a6e862f2caad4097a41dd8b260a2333e90db3b4a /heal | |
parent | 7808d4b7892ccd80bcabb1253ee359a64c51c612 (diff) |
heal: glfs-heal implementation
Thanks a lot to Niels for helping me to get build stuff right.
Change-Id: I634f24d90cd856ceab3cc0c6e9a91003f443403e
BUG: 1147462
Signed-off-by: Pranith Kumar K <pkarampu@redhat.com>
Reviewed-on: http://review.gluster.org/6529
Reviewed-by: Krutika Dhananjay <kdhananj@redhat.com>
Tested-by: Gluster Build System <jenkins@build.gluster.com>
Diffstat (limited to 'heal')
-rw-r--r-- | heal/Makefile.am | 3 | ||||
-rw-r--r-- | heal/src/Makefile.am | 35 | ||||
-rw-r--r-- | heal/src/glfs-heal.c | 610 |
3 files changed, 648 insertions, 0 deletions
diff --git a/heal/Makefile.am b/heal/Makefile.am new file mode 100644 index 00000000000..a985f42a877 --- /dev/null +++ b/heal/Makefile.am @@ -0,0 +1,3 @@ +SUBDIRS = src + +CLEANFILES = diff --git a/heal/src/Makefile.am b/heal/src/Makefile.am new file mode 100644 index 00000000000..65093df35cc --- /dev/null +++ b/heal/src/Makefile.am @@ -0,0 +1,35 @@ +sbin_PROGRAMS = glfsheal + +glfsheal_SOURCES = glfs-heal.c + +glfsheal_LDADD = $(top_builddir)/libglusterfs/src/libglusterfs.la $(GF_LDADD)\ + $(RLLIBS) $(top_builddir)/rpc/xdr/src/libgfxdr.la \ + $(top_builddir)/rpc/rpc-lib/src/libgfrpc.la \ + $(top_builddir)/api/src/libgfapi.la \ + $(top_builddir)/xlators/cluster/afr/src/afr.la \ + $(GF_GLUSTERFS_LIBS) $(XML_LIBS) + +glfsheal_LDFLAGS = $(GF_LDFLAGS) + +AM_CPPFLAGS = $(GF_CPPFLAGS) \ + -I$(top_srcdir)/xlators/lib/src\ + -I$(top_srcdir)/xlators/cluster/afr/src\ + -I$(top_srcdir)/libglusterfs/src -I$(top_srcdir)/rpc/rpc-lib/src\ + -I$(top_srcdir)/rpc/xdr/src\ + -I$(top_srcdir)/api/src\ + -I$(top_srcdir)/contrib/argp-standalone\ + -DDATADIR=\"$(localstatedir)\" \ + -DCONFDIR=\"$(sysconfdir)/glusterfs\" \ + -DGSYNCD_PREFIX=\"$(libexecdir)/glusterfs\"\ + -DSYNCDAEMON_COMPILE=$(SYNCDAEMON_COMPILE) -DSBIN_DIR=\"$(sbindir)\"\ + $(XML_CPPFLAGS) + +AM_CFLAGS = -Wall $(GF_GLUSTERFS_CFLAGS) + +CLEANFILES = + +$(top_builddir)/libglusterfs/src/libglusterfs.la: + $(MAKE) -C $(top_builddir)/libglusterfs/src/ all + +$(top_builddir)/xlators/cluster/afr/src/afr.la: + $(MAKE) -C $(top_builddir)/xlators/cluster/afr/src/ all diff --git a/heal/src/glfs-heal.c b/heal/src/glfs-heal.c new file mode 100644 index 00000000000..d02858642ab --- /dev/null +++ b/heal/src/glfs-heal.c @@ -0,0 +1,610 @@ +/* + Copyright (c) 2014 Red Hat, Inc. <http://www.redhat.com> + This file is part of GlusterFS. + + This file is licensed to you under your choice of the GNU Lesser + General Public License, version 3 or any later version (LGPLv3 or + later), or the GNU General Public License, version 2 (GPLv2), in all + cases as published by the Free Software Foundation. +*/ + +#include <stdio.h> +#include <stdlib.h> +#include <errno.h> +#include "glfs.h" +#include "glfs-handles.h" +#include "glfs-internal.h" +#include "syncop.h" +#include <string.h> +#include <time.h> +#include "afr.h" +#include "afr-self-heal.h" +#include "afr-self-heald.h" + +#define DEFAULT_HEAL_LOG_FILE_DIRECTORY DATADIR "/log/glusterfs" + +int +glfsh_link_inode_update_loc (loc_t *loc, struct iatt *iattr) +{ + inode_t *link_inode = NULL; + int ret = -1; + + link_inode = inode_link (loc->inode, NULL, NULL, iattr); + if (link_inode == NULL) + goto out; + + inode_unref (loc->inode); + loc->inode = link_inode; + ret = 0; +out: + return ret; +} + +int +glfsh_get_index_dir_loc (loc_t *rootloc, xlator_t *xl, loc_t *dirloc, + int32_t *op_errno) +{ + void *index_gfid = NULL; + int ret = 0; + dict_t *xattr = NULL; + struct iatt iattr = {0}; + struct iatt parent = {0}; + + ret = syncop_getxattr (xl, rootloc, &xattr, GF_XATTROP_INDEX_GFID); + if (ret < 0) { + *op_errno = -ret; + goto out; + } + + ret = dict_get_ptr (xattr, GF_XATTROP_INDEX_GFID, &index_gfid); + if (ret < 0) { + *op_errno = EINVAL; + goto out; + } + + uuid_copy (dirloc->gfid, index_gfid); + dirloc->path = ""; + dirloc->inode = inode_new (rootloc->inode->table); + ret = syncop_lookup (xl, dirloc, NULL, + &iattr, NULL, &parent); + dirloc->path = NULL; + if (ret < 0) { + *op_errno = -ret; + goto out; + } + ret = glfsh_link_inode_update_loc (dirloc, &iattr); + if (ret) + goto out; + glfs_loc_touchup (dirloc); + + ret = 0; +out: + if (xattr) + dict_unref (xattr); + return ret; +} + +int +afr_selfheal_locked_metadata_inspect (call_frame_t *frame, xlator_t *this, + inode_t *inode, + gf_boolean_t *metadata_selfheal) +{ + int ret = -1; + unsigned char *locked_on = NULL; + afr_private_t *priv = this->private; + + locked_on = alloca0 (priv->child_count); + + ret = afr_selfheal_inodelk (frame, this, inode, this->name, + LLONG_MAX - 1, 0, locked_on); + { + if (ret == 0) { + /* Not a single lock */ + ret = -afr_final_errno (frame->local, priv); + if (ret == 0) + ret = -ENOTCONN;/* all invalid responses */ + goto out; + } + ret = afr_selfheal_unlocked_inspect (frame, this, inode->gfid, + NULL, NULL, + metadata_selfheal, NULL); + } + afr_selfheal_uninodelk (frame, this, inode, this->name, + LLONG_MAX - 1, 0, locked_on); +out: + return ret; +} + +int +afr_selfheal_locked_data_inspect (call_frame_t *frame, xlator_t *this, + inode_t *inode, gf_boolean_t *data_selfheal) +{ + int ret = -1; + afr_private_t *priv = NULL; + unsigned char *locked_on = NULL; + unsigned char *data_lock = NULL; + + priv = this->private; + locked_on = alloca0 (priv->child_count); + data_lock = alloca0 (priv->child_count); + + ret = afr_selfheal_tryinodelk (frame, this, inode, priv->sh_domain, + 0, 0, locked_on); + { + if (ret == 0) { + ret = -afr_final_errno (frame->local, priv); + if (ret == 0) + ret = -ENOTCONN;/* all invalid responses */ + goto out; + } + ret = afr_selfheal_inodelk (frame, this, inode, this->name, + 0, 0, data_lock); + { + if (ret == 0) { + ret = -afr_final_errno (frame->local, priv); + if (ret == 0) + ret = -ENOTCONN; + /* all invalid responses */ + goto unlock; + } + ret = afr_selfheal_unlocked_inspect (frame, this, + inode->gfid, NULL, + data_selfheal, + NULL, NULL); + } + afr_selfheal_uninodelk (frame, this, inode, this->name, 0, 0, + data_lock); + } +unlock: + afr_selfheal_uninodelk (frame, this, inode, priv->sh_domain, 0, 0, + locked_on); +out: + return ret; +} + +int +afr_selfheal_locked_entry_inspect (call_frame_t *frame, xlator_t *this, + inode_t *inode, gf_boolean_t *entry_selfheal) +{ + int ret = -1; + afr_private_t *priv = NULL; + unsigned char *locked_on = NULL; + unsigned char *data_lock = NULL; + + priv = this->private; + locked_on = alloca0 (priv->child_count); + data_lock = alloca0 (priv->child_count); + + ret = afr_selfheal_tryentrylk (frame, this, inode, priv->sh_domain, + NULL, locked_on); + { + if (ret == 0) { + ret = -afr_final_errno (frame->local, priv); + if (ret == 0) + ret = -ENOTCONN;/* all invalid responses */ + goto out; + } + + ret = afr_selfheal_entrylk (frame, this, inode, this->name, + NULL, data_lock); + { + if (ret == 0) { + ret = -afr_final_errno (frame->local, priv); + if (ret == 0) + ret = -ENOTCONN; + /* all invalid responses */ + goto unlock; + } + ret = afr_selfheal_unlocked_inspect (frame, this, + inode->gfid, + NULL, NULL, NULL, + entry_selfheal); + } + afr_selfheal_unentrylk (frame, this, inode, this->name, NULL, + data_lock); + } +unlock: + afr_selfheal_unentrylk (frame, this, inode, priv->sh_domain, NULL, + locked_on); +out: + return ret; +} + +int +afr_selfheal_locked_inspect (call_frame_t *frame, xlator_t *this, uuid_t gfid, + inode_t **inode, + gf_boolean_t *data_selfheal, + gf_boolean_t *metadata_selfheal, + gf_boolean_t *entry_selfheal) +{ + int ret = -1; + gf_boolean_t dsh = _gf_false; + gf_boolean_t msh = _gf_false; + gf_boolean_t esh = _gf_false; + + ret = afr_selfheal_unlocked_inspect (frame, this, gfid, inode, + &dsh, &msh, &esh); + if (ret) + goto out; + + /* For every heal type hold locks and check if it indeed needs heal */ + + if (msh) { + msh = _gf_false; + ret = afr_selfheal_locked_metadata_inspect (frame, this, + *inode, &msh); + if (msh || ret < 0) + goto out; + } + + if (dsh) { + dsh = _gf_false; + ret = afr_selfheal_locked_data_inspect (frame, this, *inode, + &dsh); + if (dsh || ret < 0) + goto out; + } + + if (esh) { + esh = _gf_false; + ret = afr_selfheal_locked_entry_inspect (frame, this, *inode, + &esh); + } + +out: + if (entry_selfheal) + *entry_selfheal = esh; + if (data_selfheal) + *data_selfheal = dsh; + if (metadata_selfheal) + *metadata_selfheal = msh; + return ret; +} + +static xlator_t* +_get_afr_ancestor (xlator_t *xl) +{ + if (!xl || !xl->parents) + return NULL; + + while (xl->parents) { + xl = xl->parents->xlator; + if (!xl) + break; + if (strcmp (xl->type, "cluster/replicate") == 0) + return xl; + } + + return NULL; +} + +static int +glfsh_process_entries (xlator_t *xl, fd_t *fd, gf_dirent_t *entries, + off_t *offset, uint64_t *num_entries) +{ + gf_dirent_t *entry = NULL; + gf_dirent_t *tmp = NULL; + int ret = 0; + char *path = NULL; + uuid_t gfid = {0}; + inode_t *inode = NULL; + call_frame_t *frame = NULL; + xlator_t *this = NULL; + gf_boolean_t data_selfheal = _gf_false; + gf_boolean_t metadata_selfheal = _gf_false; + gf_boolean_t entry_selfheal = _gf_false; + + this = THIS; + frame = afr_frame_create (this); + if (!frame) { + ret = -1; + goto out; + } + + list_for_each_entry_safe (entry, tmp, &entries->list, list) { + *offset = entry->d_off; + if ((strcmp (entry->d_name, ".") == 0) || + (strcmp (entry->d_name, "..") == 0)) + continue; + + uuid_clear (gfid); + GF_FREE (path); + path = NULL; + if (inode) { + inode_forget (inode, 1); + inode_unref (inode); + inode = NULL; + } + + AFR_STACK_RESET (frame); + uuid_parse (entry->d_name, gfid); + ret = afr_selfheal_locked_inspect (frame, this, gfid, &inode, + &data_selfheal, + &metadata_selfheal, + &entry_selfheal); + if (ret == 0) { + if (!entry_selfheal && !metadata_selfheal && + !data_selfheal) + continue; + } + + ret = afr_shd_gfid_to_path (this, xl, gfid, &path); + + if (ret == -ENOENT || ret == -ESTALE) { + afr_shd_index_purge (xl, fd->inode, entry->d_name); + ret = 0; + continue; + } + + (*num_entries)++; + printf ("%s\n", path ? path : uuid_utoa (inode->gfid)); + } + ret = 0; +out: + if (inode) { + inode_forget (inode, 1); + inode_unref (inode); + } + + GF_FREE (path); + return ret; +} + +static int +glfsh_crawl_directory (xlator_t *readdir_xl, fd_t *fd, loc_t *loc) +{ + off_t offset = 0; + gf_dirent_t entries; + int ret = 0; + gf_boolean_t free_entries = _gf_false; + uint64_t num_entries = 0; + + INIT_LIST_HEAD (&entries.list); + + while (1) { + ret = syncop_readdir (readdir_xl, fd, 131072, offset, &entries); + if (ret <= 0) + break; + ret = 0; + free_entries = _gf_true; + + if (list_empty (&entries.list)) + goto out; + + ret = glfsh_process_entries (readdir_xl, fd, &entries, &offset, + &num_entries); + if (ret < 0) + goto out; + + gf_dirent_free (&entries); + free_entries = _gf_false; + } + ret = 0; +out: + if (free_entries) + gf_dirent_free (&entries); + if (ret < 0) { + printf ("Failed to complete gathering info. " + "Number of entries so far: %"PRIu64"\n", num_entries); + } + else { + printf ("Number of entries: %"PRIu64"\n", num_entries); + } + return ret; +} + +static int +glfsh_print_brick (xlator_t *xl, loc_t *rootloc) +{ + int ret = 0; + dict_t *xattr = NULL; + char *pathinfo = NULL; + char *brick_start = NULL; + char *brick_end = NULL; + + ret = syncop_getxattr (xl, rootloc, &xattr, GF_XATTR_PATHINFO_KEY); + if (ret < 0) + goto out; + + ret = dict_get_str (xattr, GF_XATTR_PATHINFO_KEY, &pathinfo); + if (ret < 0) + goto out; + + brick_start = strchr (pathinfo, ':') + 1; + brick_end = pathinfo + strlen (pathinfo) - 1; + *brick_end = 0; + printf ("Brick %s\n", brick_start); + +out: + if (xattr) + dict_unref (xattr); + return ret; +} + +void +glfsh_print_brick_from_xl (xlator_t *xl) +{ + char *remote_host = NULL; + char *remote_subvol = NULL; + int ret = 0; + + ret = dict_get_str (xl->options, "remote-host", &remote_host); + if (ret < 0) + goto out; + + ret = dict_get_str (xl->options, "remote-subvolume", &remote_subvol); + if (ret < 0) + goto out; +out: + if (ret < 0) + printf ("Brick - Not able to get brick information\n"); + else + printf ("Brick %s:%s\n", remote_host, remote_subvol); +} + +void +glfsh_print_pending_heals (xlator_t *xl, loc_t *rootloc) +{ + int ret = 0; + loc_t dirloc = {0}; + fd_t *fd = NULL; + int32_t op_errno = 0; + + ret = glfsh_print_brick (xl, rootloc); + if (ret < 0) { + glfsh_print_brick_from_xl (xl); + printf ("Status: %s\n", strerror (-ret)); + goto out; + } + + ret = glfsh_get_index_dir_loc (rootloc, xl, &dirloc, &op_errno); + if (ret < 0) { + if (op_errno == ESTALE || op_errno == ENOENT) + printf ("Number of entries: 0\n"); + else + printf ("Status: %s\n", strerror (op_errno)); + goto out; + } + + fd = fd_anonymous (dirloc.inode); + ret = glfsh_crawl_directory (xl, fd, &dirloc); + if (fd) + fd_unref (fd); + if (ret < 0) + printf ("Failed to find entries with pending self-heal\n"); +out: + loc_wipe (&dirloc); + return; +} + +static int +glfsh_validate_replicate_volume (xlator_t *xl) +{ + xlator_t *afr_xl = NULL; + int ret = -1; + + while (xl->next) + xl = xl->next; + + while (xl) { + if (strcmp (xl->type, "protocol/client") == 0) { + afr_xl = _get_afr_ancestor (xl); + if (afr_xl) { + ret = 0; + break; + } + } + + xl = xl->prev; + } + + return ret; +} + +int +main (int argc, char **argv) +{ + glfs_t *fs = NULL; + int ret = 0; + char *volname = NULL; + xlator_t *top_subvol = NULL; + xlator_t *xl = NULL; + loc_t rootloc = {0}; + char logfilepath[PATH_MAX] = {0}; + xlator_t *old_THIS = NULL; + xlator_t *afr_xl = NULL; + + if (argc != 2) { + printf ("Usage: %s <volname>\n", argv[0]); + ret = -1; + goto out; + } + volname = argv[1]; + + fs = glfs_new (volname); + if (!fs) { + ret = -1; + printf ("Not able to initialize volume '%s'\n", volname); + goto out; + } + + ret = glfs_set_volfile_server (fs, "tcp", "localhost", 24007); + snprintf (logfilepath, sizeof (logfilepath), + DEFAULT_HEAL_LOG_FILE_DIRECTORY"/glfsheal-%s.log", volname); + ret = glfs_set_logging(fs, logfilepath, GF_LOG_INFO); + if (ret < 0) { + ret = -1; + printf ("Not able to initialize volume '%s'\n", volname); + goto out; + } + + ret = glfs_init (fs); + if (ret < 0) { + ret = -1; + if (errno == ENOENT) { + printf ("Volume %s does not exist\n", volname); + } + else { + printf ("%s: Not able to fetch volfile from " + "glusterd\n", volname); + } + goto out; + } + + sleep (2); + __glfs_entry_fs (fs); + top_subvol = glfs_active_subvol (fs); + if (!top_subvol) { + ret = -1; + if (errno == ENOTCONN) { + printf ("Volume %s is not started (Or) All the bricks " + "are not running.\n", volname); + } + else { + printf ("%s: Not able to mount the volume, %s\n", + volname, strerror (errno)); + } + goto out; + } + + ret = glfsh_validate_replicate_volume (top_subvol); + if (ret < 0) { + printf ("Volume %s is not of type replicate\n", volname); + goto out; + } + rootloc.inode = inode_ref (top_subvol->itable->root); + glfs_loc_touchup (&rootloc); + + xl = top_subvol; + while (xl->next) + xl = xl->next; + + while (xl) { + if (strcmp (xl->type, "protocol/client") == 0) { + afr_xl = _get_afr_ancestor (xl); + if (afr_xl) { + old_THIS = THIS; + THIS = afr_xl; + glfsh_print_pending_heals (xl, &rootloc); + THIS = old_THIS; + printf("\n"); + } + } + + xl = xl->prev; + } + + loc_wipe (&rootloc); + glfs_subvol_done (fs, top_subvol); + glfs_fini (fs); + + return 0; +out: + if (fs && top_subvol) + glfs_subvol_done (fs, top_subvol); + loc_wipe (&rootloc); + if (fs) + glfs_fini (fs); + + return ret; +} |