LU-14172 lmv: optimize dir shard revalidate

mdt_is_remote_object() will check whether child is directory shard
if parent and child are on different MDTs, which needs to read LMV
from disk, and hurt striped directory stat performance.

This can be optimized, client can just set CROSS_REF flag to do a
cross reference getattr, which avoids lots of checks.

Lustre-change: https://review.whamcloud.com/40863/

Signed-off-by: Lai Siyao <lai.siyao@whamcloud.com>
Change-Id: Ib2d5a510b27c90a26f979f9cccfd40948e32d91a
Signed-off-by: Minh Diep <mdiep@whamcloud.com>
Reviewed-on: https://review.whamcloud.com/40875
Tested-by: jenkins <devops@whamcloud.com>
Reviewed-by: Andreas Dilger <adilger@whamcloud.com>
Tested-by: Maloo <maloo@whamcloud.com>
Reviewed-by: Yingjin Qian <qian@ddn.com>

diff --git a/lustre/include/obd.h b/lustre/include/obd.h
index f3dfd40..4356159 100644 (file)
--- a/lustre/include/obd.h
+++ b/lustre/include/obd.h
@@ -1128,7 +1128,7 @@ struct md_ops {
 
        int (*m_free_lustre_md)(struct obd_export *, struct lustre_md *);
 
-       int (*m_merge_attr)(struct obd_export *, const struct lu_fid *fid,
+       int (*m_merge_attr)(struct obd_export *,
                            const struct lmv_stripe_md *lsm,
                            struct cl_attr *attr, ldlm_blocking_callback);
 

diff --git a/lustre/include/obd_class.h b/lustre/include/obd_class.h
index 3875457..94432d7 100644 (file)
--- a/lustre/include/obd_class.h
+++ b/lustre/include/obd_class.h
@@ -1596,7 +1596,6 @@ static inline int md_free_lustre_md(struct obd_export *exp,
 }
 
 static inline int md_merge_attr(struct obd_export *exp,
-                               const struct lu_fid *fid,
                                const struct lmv_stripe_md *lsm,
                                struct cl_attr *attr,
                                ldlm_blocking_callback cb)
@@ -1607,7 +1606,7 @@ static inline int md_merge_attr(struct obd_export *exp,
        if (rc)
                return rc;
 
-       return MDP(exp->exp_obd, merge_attr)(exp, fid, lsm, attr, cb);
+       return MDP(exp->exp_obd, merge_attr)(exp, lsm, attr, cb);
 }
 
 static inline int md_setxattr(struct obd_export *exp, const struct lu_fid *fid,

diff --git a/lustre/include/uapi/linux/lustre/lustre_idl.h b/lustre/include/uapi/linux/lustre/lustre_idl.h
index 26fa74b..336f2d2 100644 (file)
--- a/lustre/include/uapi/linux/lustre/lustre_idl.h
+++ b/lustre/include/uapi/linux/lustre/lustre_idl.h
@@ -1934,6 +1934,13 @@ struct mdt_rec_setattr {
 
 enum mds_op_bias {
 /*     MDS_CHECK_SPLIT         = 1 << 0, obsolete before 2.3.58 */
+       /* used for remote object getattr/open by name: in the original
+        * getattr/open request, MDT found the object against name is on another
+        * MDT, then packed FID and LOOKUP lock in reply and returned -EREMOTE,
+        * and client knew it's a remote object, then set this flag in
+        * getattr/open request and sent to the corresponding MDT to finish
+        * getattr/open, which fetched attributes and UPDATE lock/opened file.
+        */
        MDS_CROSS_REF           = 1 << 1,
 /*     MDS_VTX_BYPASS          = 1 << 2, obsolete since 2.3.54 */
        MDS_PERM_BYPASS         = 1 << 3,

diff --git a/lustre/llite/file.c b/lustre/llite/file.c
index a9ca88a..71997a9 100644 (file)
--- a/lustre/llite/file.c
+++ b/lustre/llite/file.c
@@ -4418,7 +4418,7 @@ static int ll_merge_md_attr(struct inode *inode)
 
        LASSERT(lli->lli_lsm_md != NULL);
        down_read(&lli->lli_lsm_sem);
-       rc = md_merge_attr(ll_i2mdexp(inode), &lli->lli_fid, lli->lli_lsm_md,
+       rc = md_merge_attr(ll_i2mdexp(inode), ll_i2info(inode)->lli_lsm_md,
                           &attr, ll_md_blocking_ast);
        up_read(&lli->lli_lsm_sem);
        if (rc != 0)

diff --git a/lustre/llite/llite_lib.c b/lustre/llite/llite_lib.c
index 1e340ae..2545f89 100644 (file)
--- a/lustre/llite/llite_lib.c
+++ b/lustre/llite/llite_lib.c
@@ -1451,8 +1451,8 @@ static int ll_update_lsm_md(struct inode *inode, struct lustre_md *md)
                GOTO(unlock, rc = -ENOMEM);
  
        /* validate the lsm */
-       rc = md_merge_attr(ll_i2mdexp(inode), &lli->lli_fid, lli->lli_lsm_md,
-                          attr, ll_md_blocking_ast);
+       rc = md_merge_attr(ll_i2mdexp(inode), lli->lli_lsm_md, attr,
+                          ll_md_blocking_ast);
        if (!rc) {
                if (md->body->mbo_valid & OBD_MD_FLNLINK)
                        md->body->mbo_nlink = attr->cat_nlink;

diff --git a/lustre/lmv/lmv_intent.c b/lustre/lmv/lmv_intent.c
index c3e2ad7..361d970 100644 (file)
--- a/lustre/lmv/lmv_intent.c
+++ b/lustre/lmv/lmv_intent.c
@@ -141,7 +141,6 @@ out:
 }
 
 int lmv_revalidate_slaves(struct obd_export *exp,
-                         const struct lu_fid *pfid,
                          const struct lmv_stripe_md *lsm,
                          ldlm_blocking_callback cb_blocking,
                          int extra_lock_flags)
@@ -188,11 +187,14 @@ int lmv_revalidate_slaves(struct obd_export *exp,
                 * which is not needed here.
                 */
                memset(op_data, 0, sizeof(*op_data));
-               if (exp_connect_flags2(exp) & OBD_CONNECT2_GETATTR_PFID)
-                       op_data->op_fid1 = *pfid;
-               else
-                       op_data->op_fid1 = fid;
+               op_data->op_fid1 = fid;
                op_data->op_fid2 = fid;
+               /* shard revalidate only needs to fetch attributes and UPDATE
+                * lock, which is similar to the bottom half of remote object
+                * getattr, set this flag so that MDT skips checking whether
+                * it's remote object.
+                */
+               op_data->op_bias = MDS_CROSS_REF;
 
                tgt = lmv_get_target(lmv, lsm->lsm_md_oinfo[i].lmo_mds, NULL);
                if (IS_ERR(tgt))
@@ -474,8 +476,7 @@ retry:
                /* If RPC happens, lsm information will be revalidated
                 * during update_inode process (see ll_update_lsm_md) */
                if (op_data->op_mea2 != NULL) {
-                       rc = lmv_revalidate_slaves(exp, &op_data->op_fid2,
-                                                  op_data->op_mea2,
+                       rc = lmv_revalidate_slaves(exp, op_data->op_mea2,
                                                   cb_blocking,
                                                   extra_lock_flags);
                        if (rc != 0)

diff --git a/lustre/lmv/lmv_internal.h b/lustre/lmv/lmv_internal.h
index 72777c3..0ad7432 100644 (file)
--- a/lustre/lmv/lmv_internal.h
+++ b/lustre/lmv/lmv_internal.h
@@ -54,7 +54,6 @@ int lmv_fid_alloc(const struct lu_env *env, struct obd_export *exp,
                  struct lu_fid *fid, struct md_op_data *op_data);
 
 int lmv_revalidate_slaves(struct obd_export *exp,
-                         const struct lu_fid *pfid,
                          const struct lmv_stripe_md *lsm,
                          ldlm_blocking_callback cb_blocking,
                          int extra_lock_flags);

diff --git a/lustre/lmv/lmv_obd.c b/lustre/lmv/lmv_obd.c
index b94fa61..078f6e2 100644 (file)
--- a/lustre/lmv/lmv_obd.c
+++ b/lustre/lmv/lmv_obd.c
@@ -3453,7 +3453,6 @@ int lmv_quotactl(struct obd_device *unused, struct obd_export *exp,
 }
 
 static int lmv_merge_attr(struct obd_export *exp,
-                         const struct lu_fid *fid,
                          const struct lmv_stripe_md *lsm,
                          struct cl_attr *attr,
                          ldlm_blocking_callback cb_blocking)
@@ -3461,7 +3460,7 @@ static int lmv_merge_attr(struct obd_export *exp,
        int rc;
        int i;
 
-       rc = lmv_revalidate_slaves(exp, fid, lsm, cb_blocking, 0);
+       rc = lmv_revalidate_slaves(exp, lsm, cb_blocking, 0);
        if (rc < 0)
                return rc;
 

diff --git a/lustre/mdt/mdt_lib.c b/lustre/mdt/mdt_lib.c
index bf4403f..2818339 100644 (file)
--- a/lustre/mdt/mdt_lib.c
+++ b/lustre/mdt/mdt_lib.c
@@ -1784,29 +1784,6 @@ static int mdt_fids_different_target(struct mdt_thread_info *info,
        return index1 != index2;
 }
 
-static bool mdt_object_is_shard(struct mdt_thread_info *info,
-                               struct mdt_object *obj)
-{
-       struct lmv_mds_md_v1 *lmv = (struct lmv_mds_md_v1 *)info->mti_xattr_buf;
-       struct lu_buf buf;
-       int rc;
-
-       if (!mdt_object_exists(obj))
-               return false;
-
-       if (!S_ISDIR(lu_object_attr(&obj->mot_obj)))
-               return false;
-
-       buf.lb_buf = lmv;
-       buf.lb_len = sizeof(*lmv);
-       rc = mo_xattr_get(info->mti_env, mdt_object_child(obj), &buf,
-                         XATTR_NAME_LMV);
-       if (rc < 0)
-               return false;
-
-       return lmv->lmv_magic == cpu_to_le32(LMV_MAGIC_STRIPE);
-}
-
 /**
  * Check whether \a child is remote object on \a parent.
  *
@@ -1839,14 +1816,8 @@ int mdt_is_remote_object(struct mdt_thread_info *info,
                RETURN(0);
 
        if (likely(parent != child)) {
-               if (mdt_object_remote(parent) ^ mdt_object_remote(child)) {
-                       /* don't treat shard as remote object, otherwise client
-                        * need to revalidate shards all the time.
-                        */
-                       if (mdt_object_is_shard(info, child))
-                               RETURN(0);
+               if (mdt_object_remote(parent) ^ mdt_object_remote(child))
                        RETURN(1);
-               }
 
                if (!mdt_object_remote(parent) && !mdt_object_remote(child))
                        RETURN(0);