Viewing: lod_qos.c
// SPDX-License-Identifier: GPL-2.0
/*
* Copyright 2009 Sun Microsystems, Inc. All rights reserved
* Use is subject to license terms.
*
* Copyright (c) 2012, 2017, Intel Corporation.
*/
/*
* This file is part of Lustre, http://www.lustre.org/
*
* Implementation of different allocation algorithm used
* to distribute objects and data among OSTs.
*/
#define DEBUG_SUBSYSTEM S_LOV
#include <asm/div64.h>
#include <linux/random.h>
#include <lustre_compat/linux/timer.h>
#include <uapi/linux/lustre/lustre_idl.h>
#include <lustre_swab.h>
#include <obd_class.h>
#include "lod_internal.h"
/* check whether a target is available for new object allocation */
static inline int lod_statfs_check(struct lu_tgt_descs *ltd,
struct lu_tgt_desc *tgt)
{
struct obd_statfs *sfs = &tgt->ltd_statfs;
if (sfs->os_state & OS_STATFS_ENOSPC ||
(sfs->os_state & OS_STATFS_ENOINO &&
/* OST allocation allowed while precreated objects available */
(ltd->ltd_is_mdt || sfs->os_fprecreated == 0)))
return -ENOSPC;
/* If the OST is readonly then we can't allocate objects there */
if (sfs->os_state & OS_STATFS_READONLY)
return -EROFS;
/* object creation is skipped on the OST with max_create_count=0 */
if (!ltd->ltd_is_mdt && sfs->os_state & OS_STATFS_NOCREATE)
return -ENOBUFS;
return 0;
}
/**
* lod_statfs_and_check() - Check whether target is available for new objects.
* @env: execution environment for this thread
* @d: LOD device
* @ltd: target table
* @tgt: target
* @reserve: space to reserve on target device
*
* Request statfs data from the given target and verify it's active and not
* read-only. If so, then it can be used to place new objects. This
* function also maintains the number of active/inactive targets and sets
* dirty flags if those numbers change so others can run re-balance procedures.
* No external locking is required.
*
* Return:
* * %0 if the target is good
* * %negative negated errno on error
*/
static int lod_statfs_and_check(const struct lu_env *env, struct lod_device *d,
struct lu_tgt_descs *ltd,
struct lu_tgt_desc *tgt, __u64 reserve)
{
struct obd_statfs_info info = { 0 };
struct lov_desc *desc = <d->ltd_lov_desc;
int rc;
ENTRY;
LASSERT(d);
LASSERT(tgt);
info.os_enable_pre = 1;
rc = dt_statfs_info(env, tgt->ltd_tgt, &tgt->ltd_statfs, &info);
if (rc && rc != -ENOTCONN)
CERROR("%s: statfs error: rc = %d\n", lod2obd(d)->obd_name, rc);
if (!rc) {
tgt->ltd_qos.ltq_failure_domain = tgt->ltd_statfs.os_failure_domain;
rc = lod_statfs_check(ltd, tgt);
}
/* reserving space shouldn't be enough to mark an OST inactive */
if (reserve &&
(reserve + (info.os_reserved_mb_low << 20) >
tgt->ltd_statfs.os_bavail * tgt->ltd_statfs.os_bsize))
return -ENOSPC;
/* check whether device has changed state (active, inactive) */
if (rc && tgt->ltd_active) {
/* turned inactive? */
spin_lock(&d->lod_lock);
if (tgt->ltd_active) {
tgt->ltd_active = 0;
if (rc == -ENOTCONN)
tgt->ltd_discon = 1;
LASSERT(desc->ld_active_tgt_count > 0);
desc->ld_active_tgt_count--;
set_bit(LQ_DIRTY, <d->ltd_qos.lq_flags);
CDEBUG(D_CONFIG, "%s: turns inactive\n",
tgt->ltd_exp->exp_obd->obd_name);
}
spin_unlock(&d->lod_lock);
} else if (rc == 0 && !tgt->ltd_active) {
/* turned active? */
spin_lock(&d->lod_lock);
if (!tgt->ltd_active) {
LASSERTF(desc->ld_active_tgt_count < desc->ld_tgt_count,
"active tgt count %d, tgt nr %d\n",
desc->ld_active_tgt_count, desc->ld_tgt_count);
tgt->ltd_active = 1;
tgt->ltd_discon = 0;
desc->ld_active_tgt_count++;
set_bit(LQ_DIRTY, <d->ltd_qos.lq_flags);
CDEBUG(D_CONFIG, "%s: turns active\n",
tgt->ltd_exp->exp_obd->obd_name);
}
spin_unlock(&d->lod_lock);
}
if (rc == -ENOTCONN) {
/* In case that the ENOTCONN for inactive OST state is
* mistreated as MDT disconnection state by the client,
* this error should be changed to someone else.
*/
rc = -EREMOTEIO;
}
RETURN(rc);
}
/**
* lod_qos_statfs_update() - Maintain per-target statfs data.
* @env: execution environment for this thread
* @lod: LOD device
* @ltd: tgt table
*
* The function refreshes statfs data for all the targets every N seconds.
* The actual N is controlled via procfs and set to LOV_DESC_QOS_MAXAGE_DEFAULT
* initially.
*/
void lod_qos_statfs_update(const struct lu_env *env, struct lod_device *lod,
struct lu_tgt_descs *ltd)
{
struct obd_device *obd = lod2obd(lod);
struct lu_tgt_desc *tgt;
time64_t max_age;
u64 avail;
ENTRY;
max_age = ktime_get_seconds() - 2 * ltd->ltd_lov_desc.ld_qos_maxage;
if (obd->obd_osfs_age > max_age)
/* statfs data are quite recent, don't need to refresh it */
RETURN_EXIT;
if (test_and_set_bit(LQ_SF_PROGRESS, <d->ltd_qos.lq_flags))
RETURN_EXIT;
if (obd->obd_osfs_age > max_age) {
/* statfs data are quite recent, don't need to refresh it */
clear_bit(LQ_SF_PROGRESS, <d->ltd_qos.lq_flags);
RETURN_EXIT;
}
lod_getref(ltd);
ltd_foreach_tgt(ltd, tgt) {
avail = tgt->ltd_statfs.os_bavail;
if (lod_statfs_and_check(env, lod, ltd, tgt, 0))
continue;
if (tgt->ltd_statfs.os_bavail != avail)
/* recalculate weigths */
set_bit(LQ_DIRTY, <d->ltd_qos.lq_flags);
}
lod_putref(lod, ltd);
obd->obd_osfs_age = ktime_get_seconds();
clear_bit(LQ_SF_PROGRESS, <d->ltd_qos.lq_flags);
EXIT;
}
#define LOV_QOS_EMPTY ((__u32)-1)
/**
* lod_qos_calc_rr() - Calculate optimal round-robin order with regard to OSSes
* @lod: LOD device
* @ltd: tgt table
* @src_pool: tgt pool
* @lqr: round-robin list
*
* Place all the OSTs from pool @src_pool in a special array to be used for
* round-robin (RR) stripe allocation. The placement algorithm interleaves
* OSTs from the different OSSs so that RR allocation can balance OSSs evenly.
* Resorts the targets when the number of active targets changes (because of
* a new target or activation/deactivation).
*
* Return:
* * %0 on success
* * %-ENOMEM fails to allocate the array
*/
static int lod_qos_calc_rr(struct lod_device *lod, struct lu_tgt_descs *ltd,
const struct lu_tgt_pool *src_pool,
struct lu_qos_rr *lqr)
{
struct lu_svr_qos *svr;
struct lu_tgt_desc *tgt;
unsigned placed, real_count;
unsigned int i;
int rc;
ENTRY;
if (!test_bit(LQ_DIRTY, &lqr->lqr_flags)) {
LASSERT(lqr->lqr_pool.op_size);
RETURN(0);
}
/* Do actual allocation. */
down_write(<d->ltd_qos.lq_rw_sem);
/*
* Check again. While we were sleeping on @lq_rw_sem something could
* change.
*/
if (!test_bit(LQ_DIRTY, &lqr->lqr_flags)) {
LASSERT(lqr->lqr_pool.op_size);
up_write(<d->ltd_qos.lq_rw_sem);
RETURN(0);
}
real_count = src_pool->op_count;
/* Zero the pool array */
/* alloc_rr is holding a read lock on the pool, so nobody is adding/
deleting from the pool. The lq_rw_sem insures that nobody else
is reading. */
lqr->lqr_pool.op_count = real_count;
rc = lu_tgt_pool_extend(&lqr->lqr_pool, real_count);
if (rc) {
up_write(<d->ltd_qos.lq_rw_sem);
RETURN(rc);
}
for (i = 0; i < lqr->lqr_pool.op_count; i++)
lqr->lqr_pool.op_array[i] = LOV_QOS_EMPTY;
/* Place all the tgts from 1 svr at the same time. */
placed = 0;
list_for_each_entry(svr, <d->ltd_qos.lq_svr_list, lsq_svr_list) {
int j = 0;
for (i = 0; i < lqr->lqr_pool.op_count; i++) {
int next;
if (!test_bit(src_pool->op_array[i],
ltd->ltd_tgt_bitmap))
continue;
tgt = LTD_TGT(ltd, src_pool->op_array[i]);
LASSERT(tgt && tgt->ltd_tgt);
if (tgt->ltd_qos.ltq_svr != svr)
continue;
/* Evenly space these tgts across arrayspace */
next = j * lqr->lqr_pool.op_count / svr->lsq_tgt_count;
while (lqr->lqr_pool.op_array[next] != LOV_QOS_EMPTY)
next = (next + 1) % lqr->lqr_pool.op_count;
lqr->lqr_pool.op_array[next] = src_pool->op_array[i];
j++;
placed++;
}
}
clear_bit(LQ_DIRTY, &lqr->lqr_flags);
up_write(<d->ltd_qos.lq_rw_sem);
if (placed != real_count) {
/* This should never happen */
LCONSOLE_ERROR("Failed to place all tgts in the round-robin list (%d of %d).\n",
placed, real_count);
for (i = 0; i < lqr->lqr_pool.op_count; i++) {
LCONSOLE(D_WARNING, "rr #%d tgt idx=%d\n", i,
lqr->lqr_pool.op_array[i]);
}
set_bit(LQ_DIRTY, &lqr->lqr_flags);
RETURN(-EAGAIN);
}
RETURN(0);
}
/**
* lod_qos_declare_object_on() - Instantiate & declare creation of a new object.
* @env: execution environment for this thread
* @d: LOD device
* @ost_idx: OST target index where the object is being created
* @can_block: operation blockable or not
* @th: transaction handle
*
* The function instantiates LU representation for a new object on the
* specified device. Also it declares an intention to create that
* object on the storage target.
*
* Note lu_object_anon() is used which is a trick with regard to LU/OSD
* infrastructure - in the existing precreation framework we can't assign FID
* at this moment, we do this later once a transaction is started. So the
* special method instantiates FID-less object in the cache and later it
* will get a FID and proper placement in LU cache.
*
* Return object ptr on success, ERR_PTR() otherwise
*/
static struct dt_object *lod_qos_declare_object_on(const struct lu_env *env,
struct lod_device *d,
__u32 ost_idx,
bool can_block,
struct thandle *th)
{
struct dt_allocation_hint *ah = &lod_env_info(env)->lti_ah;
struct lod_tgt_desc *ost;
struct lu_object *o, *n;
struct lu_device *nd;
struct dt_object *dt;
int rc;
ENTRY;
LASSERT(d);
LASSERT(ost_idx < d->lod_ost_descs.ltd_tgts_size);
ost = OST_TGT(d,ost_idx);
LASSERT(ost);
LASSERT(ost->ltd_tgt);
nd = &ost->ltd_tgt->dd_lu_dev;
/*
* allocate anonymous object with zero fid, real fid
* will be assigned by OSP within transaction
* XXX: to be fixed with fully-functional OST fids
*/
o = lu_object_anon(env, nd, NULL);
if (IS_ERR(o))
GOTO(out, dt = ERR_CAST(o));
n = lu_object_locate(o->lo_header, nd->ld_type);
if (unlikely(n == NULL)) {
CERROR("can't find slice\n");
lu_object_put(env, o);
GOTO(out, dt = ERR_PTR(-EINVAL));
}
dt = container_of(n, struct dt_object, do_lu);
ah->dah_can_block = can_block;
rc = lod_sub_declare_create(env, dt, NULL, ah, NULL, th);
if (rc < 0) {
CDEBUG(D_OTHER, "can't declare creation on #%u: %d\n",
ost_idx, rc);
lu_object_put(env, o);
dt = ERR_PTR(rc);
}
out:
RETURN(dt);
}
/**
* ec_raidset_count() - Number of raid sets in an EC split.
* @sc: split information
*
* Return: the total number of raid sets.
*/
static inline int ec_raidset_count(const struct ec_split_comp *sc)
{
return sc->esc_n0 + sc->esc_n1;
}
/**
* lod_stripe_count_min() - Calculate a minimum acceptable stripe count.
* @lod_comp: the component we are allocating
* @flags: 0 or LOD_USES_DEFAULT_STRIPE (plain components only)
* @sc: raid-set split; meaningful for EC components
* @include_cstripe: if true, add llc_cstripe_count into data-component mins
* (QoS does; RR does not)
*
* Parity components (LCME_FL_PARITY):
* - single raid set: llc_cstripe_count unique OSTs
* - multiple raid sets: 0 (parity can reuse data OSTs from other sets)
*
* EC-protected data components (llc_dstripe_count != 0):
* - single raid set: stripe_count, plus cstripe when @include_cstripe
* - multi-set with overstriping: esc_k0, plus cstripe when @include_cstripe
* - multi-set without overstriping: prefer fewer raid sets over a smaller k:
* nsets >= 4 → (nsets - nsets/4) * esc_k0
* nsets < 4 → esc_k0 - esc_k0/4
* then plus cstripe when @include_cstripe. Uses x - x/4 so values under
* 4 do not collapse to 0 the way (x * 3 / 4) would.
*
* Plain (non-EC) components:
* - LOD_USES_DEFAULT_STRIPE: stripe_count - stripe_count/4
* - otherwise: stripe_count
*
* Return: minimum number of unique OSTs the allocator must place before
* EC fallbacks (reuse/shrink) or plain partial allocation may run
*/
static int lod_stripe_count_min(struct lod_layout_component *lod_comp,
enum lod_uses_hint flags,
struct ec_split_comp *sc, bool include_cstripe)
{
__u32 stripe_count = lod_comp->llc_stripe_count;
int maybe_cstripe;
int nsets, min_stripes;
/* This is an EC components */
if (lod_comp->llc_flags & LCME_FL_PARITY) {
if (ec_raidset_count(sc) == 1)
return lod_comp->llc_cstripe_count;
return 0;
}
maybe_cstripe = include_cstripe ? lod_comp->llc_cstripe_count : 0;
/* This is a data component with a matching ec */
if (lod_comp->llc_dstripe_count) {
if (ec_raidset_count(sc) == 1)
return stripe_count + maybe_cstripe;
if (lod_comp->llc_pattern & LOV_PATTERN_OVERSTRIPING)
return sc->esc_k0 + maybe_cstripe;
/*
* Multi-raidset without overstriping: allowing a reduced
* minimum means create can still succeed (and shrink via
* lod_ec_shrink_data_geometry) when some OSTs are offline,
* e.g. --ec 6+2 on 8 OSTs with one down. Prefer fewer raid
* sets over fewer stripes per set.
*
* nsets >= 4: keep at least 3/4 of the raid sets (at esc_k0)
* nsets < 4: keep at least 3/4 of esc_k0
* (stripe - stripe/4 avoids *3/4 == 0 when value < 4)
*/
nsets = ec_raidset_count(sc);
if (nsets >= 4)
min_stripes = (nsets - nsets / 4) * sc->esc_k0;
else
min_stripes = sc->esc_k0 - sc->esc_k0 / 4;
if (min_stripes > stripe_count)
min_stripes = stripe_count;
if (min_stripes < 1)
min_stripes = 1;
return min_stripes + maybe_cstripe;
}
return (flags & LOD_USES_DEFAULT_STRIPE ?
stripe_count - (stripe_count / 4) : stripe_count);
}
#define LOV_CREATE_RESEED_MULT 30
#define LOV_CREATE_RESEED_MIN 2000
/**
* lod_qos_tgt_in_use_clear() - Initialize temporary tgt-in-use array.
* @env: execution environment for this thread
* @stripes: number of items needed in the array
*
* Allocate or extend the array used to mark targets already assigned to a new
* striping so they are not used more than once.
*
* Return:
* * %0 on success
* * %-ENOMEM on error
*/
static inline int lod_qos_tgt_in_use_clear(const struct lu_env *env,
__u32 stripes)
{
struct lod_thread_info *info = lod_env_info(env);
if (info->lti_ea_buf.lb_len < sizeof(int) * stripes)
lod_ea_store_resize(info, stripes * sizeof(int));
if (info->lti_ea_buf.lb_len < sizeof(int) * stripes) {
CERROR("can't allocate memory for tgt-in-use array\n");
return -ENOMEM;
}
memset(info->lti_ea_store, -1, sizeof(int) * stripes);
return 0;
}
/**
* lod_qos_tgt_in_use() - Remember a target in the array of used targets.
* @env: execution environment for this thread
* @idx: index in the array
* @tgt_idx: target index to mark as used
*
* Mark the given target as used for a new striping being created. The status
* of an tgt in a striping can be checked with lod_qos_is_tgt_used().
*/
static inline void lod_qos_tgt_in_use(const struct lu_env *env,
int idx, int tgt_idx)
{
struct lod_thread_info *info = lod_env_info(env);
int *tgts = info->lti_ea_buf.lb_buf;
LASSERT(info->lti_ea_buf.lb_len >= idx * sizeof(int));
tgts[idx] = tgt_idx;
}
/**
* lod_qos_is_tgt_used() - Check is tgt used in a striping.
* @env: execution environment for this thread
* @tgt_idx: target index to check
* @stripes: the number of items used in the array already
*
* Checks whether tgt with the given index is marked as used in the temporary
* array (see lod_qos_tgt_in_use()).
*
* Return:
* * %0 not used
* * %1 used
*/
static int lod_qos_is_tgt_used(const struct lu_env *env, int tgt_idx,
__u32 stripes)
{
struct lod_thread_info *info = lod_env_info(env);
int *tgts = info->lti_ea_buf.lb_buf;
__u32 j;
for (j = 0; j < stripes; j++) {
if (tgts[j] == tgt_idx)
return 1;
}
return 0;
}
static inline bool
lod_obj_is_ost_use_skip_cb(const struct lu_env *env, struct lod_object *lo,
int comp_idx, struct lod_obj_stripe_cb_data *data)
{
struct lod_layout_component *comp = &lo->ldo_comp_entries[comp_idx];
return comp->llc_ost_indices == NULL;
}
static inline int
lod_obj_is_ost_use_cb(const struct lu_env *env, struct lod_object *lo,
int comp_idx, struct lod_obj_stripe_cb_data *data)
{
struct lod_layout_component *comp = &lo->ldo_comp_entries[comp_idx];
int i;
for (i = 0; i < comp->llc_stripe_count; i++) {
if (comp->llc_ost_indices[i] == data->locd_ost_index) {
data->locd_ost_index = -1;
return -EEXIST;
}
}
return 0;
}
/**
* lod_comp_is_ost_used() - Check is OST used in a composite layout
* @env: execution environment
* @lo: lod object
* @ost: OST target index to check
*
* Return:
* * %false not used
* * %true used
*/
static inline bool lod_comp_is_ost_used(const struct lu_env *env,
struct lod_object *lo, int ost)
{
struct lod_obj_stripe_cb_data data = { { 0 } };
data.locd_ost_index = ost;
data.locd_comp_skip_cb = lod_obj_is_ost_use_skip_cb;
data.locd_comp_cb = lod_obj_is_ost_use_cb;
(void)lod_obj_for_each_stripe(env, lo, NULL, &data);
return data.locd_ost_index == -1;
}
static inline void lod_avoid_update(struct lod_object *lo,
struct lod_avoid_guide *lag)
{
if (!lod_is_flr(lo))
return;
lag->lag_ost_avail--;
}
static inline bool lod_should_avoid_ost(struct lod_object *lo,
struct lod_avoid_guide *lag,
__u32 index)
{
struct lod_device *lod = lu2lod_dev(lo->ldo_obj.do_lu.lo_dev);
struct lod_tgt_desc *ost = OST_TGT(lod, index);
struct lu_svr_qos *lsq = ost->ltd_qos.ltq_svr;
bool used = false;
int i;
if (!test_bit(index, lod->lod_ost_bitmap)) {
CDEBUG(D_OTHER, "OST%d: been used in conflicting mirror component\n",
index);
return true;
}
/**
* we've tried our best, all available OSTs have been used in
* overlapped components in the other mirror
*/
if (lag->lag_ost_avail == 0)
return false;
/* check OSS use */
for (i = 0; i < lag->lag_oaa_count; i++) {
if (lag->lag_oss_avoid_array[i] == lsq->lsq_id) {
used = true;
break;
}
}
/**
* if the OSS which OST[index] resides has not been used, we'd like to
* use it
*/
if (!used)
return false;
/* if the OSS has been used, check whether the OST has been used */
if (!test_bit(index, lag->lag_ost_avoid_bitmap))
used = false;
else
CDEBUG(D_OTHER, "OST%d: been used in conflicting mirror component\n",
index);
return used;
}
static int lod_check_and_reserve_ost(const struct lu_env *env,
struct lod_object *lo,
struct lod_layout_component *lod_comp,
__u32 ost_idx, __u32 speed, __u32 *s_idx,
struct dt_object **stripe,
__u32 *ost_indices,
struct thandle *th,
bool *overstriped,
__u64 reserve)
{
struct lod_device *lod = lu2lod_dev(lo->ldo_obj.do_lu.lo_dev);
struct lod_avoid_guide *lag = &lod_env_info(env)->lti_avoid;
struct lu_tgt_desc *ost = OST_TGT(lod, ost_idx);
struct dt_object *o;
__u32 stripe_idx = *s_idx;
int rc;
ENTRY;
rc = lod_statfs_and_check(env, lod, &lod->lod_ost_descs, ost, reserve);
if (rc)
RETURN(rc);
/*
* We expect number of precreated objects in f_ffree at
* the first iteration, skip OSPs with no objects ready
*/
if (ost->ltd_statfs.os_fprecreated == 0 && speed == 0) {
CDEBUG(D_OTHER, "#%d: precreation is empty\n", ost_idx);
RETURN(rc);
}
/*
* try to use another OSP if this one is degraded
*/
if (ost->ltd_statfs.os_state & OS_STATFS_DEGRADED && speed < 2) {
CDEBUG(D_OTHER, "#%d: degraded\n", ost_idx);
RETURN(rc);
}
if (lod_comp->llc_cstripe_count &&
(lag->lag_ost_avail <= 0 ||
lod_should_avoid_ost(lo, lag, ost_idx))) {
CDEBUG(D_OTHER,
"Skip conflicting OST%d because we want unique OSTs\n",
ost_idx);
RETURN(rc);
}
/*
* try not allocate on OST which has been used by other
* component
*/
if (speed == 0 && lod_comp_is_ost_used(env, lo, ost_idx)) {
CDEBUG(D_OTHER, "iter %d: OST%d used by other component\n",
speed, ost_idx);
RETURN(rc);
}
/**
* try not allocate OSTs used by conflicting component of other mirrors
* for the first and second time.
*/
if (speed < 2 && lod_should_avoid_ost(lo, lag, ost_idx)) {
CDEBUG(D_OTHER, "iter %d: OST%d used by conflicting mirror component\n",
speed, ost_idx);
RETURN(rc);
}
/* do not put >1 objects on a single OST, except for overstriping */
if (lod_qos_is_tgt_used(env, ost_idx, stripe_idx)) {
if (lod_comp->llc_pattern & LOV_PATTERN_OVERSTRIPING)
*overstriped = true;
else
RETURN(rc);
}
o = lod_qos_declare_object_on(env, lod, ost_idx, (speed > 1), th);
if (IS_ERR(o)) {
CDEBUG(D_OTHER, "can't declare new object on #%u: %d\n",
ost_idx, (int) PTR_ERR(o));
rc = PTR_ERR(o);
RETURN(rc);
}
/*
* We've successfully declared (reserved) an object
*/
lod_avoid_update(lo, lag);
lod_qos_tgt_in_use(env, stripe_idx, ost_idx);
stripe[stripe_idx] = o;
ost_indices[stripe_idx] = ost_idx;
CFS_FAIL_TIMEOUT(OBD_FAIL_MDS_LOV_CREATE_RACE, 2);
stripe_idx++;
*s_idx = stripe_idx;
RETURN(rc);
}
/**
* lod_ec_add_parities() - Add additional parity OSTs to parity_comp.
* @env: execution environment for this thread
* @lo: LOD object
* @th: transaction handle
* @nfound: current number of found/selected OSTs in parity_comp
* @sc: split information
* @data_comp: the data component which has already been allocated OSTs
* @parity_comp: the current parity comp.
* @stripe: striping created
* @ost_indices: ost indices of striping created
*
* This function is used to fill the remaining OSTs for the parity
* component, parity_comp, by reusing OSTs from the associated data component.
*/
static __u32 lod_ec_add_parities(const struct lu_env *env,
struct lod_object *lo, struct thandle *th,
__u32 nfound, struct ec_split_comp *sc,
struct lod_layout_component *data_comp,
struct lod_layout_component *parity_comp,
struct dt_object **stripe, __u32 *ost_indices)
{
int i, pos, tries;
struct lod_device *lod = lu2lod_dev(lo->ldo_obj.do_lu.lo_dev);
struct dt_object *o;
int cur_rs;
if (!data_comp->llc_stripes_allocated ||
!parity_comp->llc_cstripe_count)
return nfound;
for (pos = 0, tries = 0;
nfound < parity_comp->llc_stripe_count &&
tries <= data_comp->llc_stripe_count;
pos = (pos + 1) % data_comp->llc_stripe_count, tries++) {
bool already_used;
int start, num;
/* Current raid set we are trying to assign parities to */
cur_rs = nfound / parity_comp->llc_cstripe_count;
/* make sure this OST is not part of current raid set */
if (cur_rs < sc->esc_n0) {
start = cur_rs * sc->esc_k0;
num = sc->esc_k0;
} else {
start = (sc->esc_n0 * sc->esc_k0) +
(cur_rs - sc->esc_n0) * sc->esc_k1;
num = sc->esc_k1;
}
already_used = false;
for (i = start; i < start + num; i++) {
if (data_comp->llc_ost_indices[i] ==
data_comp->llc_ost_indices[pos]) {
already_used = true;
break;
}
}
if (already_used)
continue;
/* We can't use it if we already used it in this ec raid set
* either
*/
for (i = cur_rs * parity_comp->llc_cstripe_count;
i < nfound; i++) {
if (ost_indices[i] ==
data_comp->llc_ost_indices[pos]) {
already_used = true;
break;
}
}
if (already_used)
continue;
o = lod_qos_declare_object_on(env, lod,
data_comp->llc_ost_indices[pos], true, th);
if (IS_ERR(o)) {
CDEBUG(D_OTHER, "can't declare object on #%u: %d\n",
data_comp->llc_ost_indices[pos],
(int) PTR_ERR(o));
return nfound;
}
stripe[nfound] = o;
ost_indices[nfound] = data_comp->llc_ost_indices[pos];
nfound++;
/* Found an OST so reset the tries counter to 0 */
tries = 0;
}
return nfound;
}
/**
* lod_ec_check_data_usable() - Checks if a OST conflicts with data comp
* @idx: index of OST candidate
* @nfound: current number of found/selected OSTs in lod_comp
* @sc: split information
* @data_comp: the data component which has already been allocated OSTs
* @parity_comp: the current parity comp.
*
* This function is used when trying to add stripes to a parity component.
* It verifies that the current candidate idx is not already used in the
* corresponding data stripe set / raid set.
*/
static bool lod_ec_check_data_usable(__u32 idx, __u32 nfound,
struct ec_split_comp *sc,
struct lod_layout_component *data_comp,
struct lod_layout_component *parity_comp)
{
int start, num;
int i, cur_rs;
if (!data_comp->llc_stripes_allocated ||
!parity_comp->llc_cstripe_count)
return true;
if (parity_comp->llc_pattern & LOV_PATTERN_OVERSTRIPING) {
/* Just check the data OSTs in the current raidset */
cur_rs = nfound / parity_comp->llc_cstripe_count;
/* make sure this OST is not part of current raid set */
if (cur_rs < sc->esc_n0) {
start = cur_rs * sc->esc_k0;
num = sc->esc_k0;
} else {
start = (sc->esc_n0 * sc->esc_k0) +
(cur_rs - sc->esc_n0) * sc->esc_k1;
num = sc->esc_k1;
}
} else {
/* Can not reuse any OST from data comp */
start = 0;
num = data_comp->llc_stripe_count;
}
for (i = start; i < start + num; i++)
if (data_comp->llc_ost_indices[i] == idx)
return false;
return true;
}
/**
* lod_ec_shrink_data_geometry() - Shrink EC data layout to @nfound OSTs.
* @nfound: number of unique OSTs already allocated
* @data_comp: EC-protected data component being allocated
* @sc: raid-set split describing the current layout (updated in place)
*
* When not enough unique OSTs are available for the requested data stripe
* count, convert esc_k0 raidsets into esc_k1 raidsets one stripe at a time
* until the geometry fits in @nfound OSTs. This avoids failing create when
* some OSTs are offline for a multi-raidset EC layout without overstriping.
*
* Objects already placed in stripe[0..nfound) form a dense prefix of the
* layout, so after reducing the total stripe count no raidset reorder is
* required — esc_k1 sets naturally sit at the end of the resplit geometry.
*
* Return: new stripe count (== @nfound) if the layout can be shrunk to fit,
* otherwise the original stripe count (caller should still fail).
*/
static __u32 lod_ec_shrink_data_geometry(__u32 nfound,
struct lod_layout_component *data_comp,
struct ec_split_comp *sc)
{
__u32 stripe_count = data_comp->llc_stripe_count;
__u32 suggested = data_comp->llc_dstripe_count;
__u32 orig = stripe_count;
if (nfound == 0 || nfound >= stripe_count || suggested == 0)
return stripe_count;
/*
* Convert one esc_k0 raidset to an esc_k1 raidset per step
* (drop one data stripe). Standard split has esc_k0 == esc_k1 + 1
* when both sizes are present; a uniform layout has esc_n1 == 0.
*/
while (stripe_count > nfound && sc->esc_n0 > 0 && sc->esc_k0 > 1) {
if (sc->esc_n1 == 0) {
/* Uniform k0 sets → first k1 set of size k0 - 1 */
sc->esc_k1 = sc->esc_k0 - 1;
sc->esc_n0--;
sc->esc_n1 = 1;
} else if (sc->esc_k1 == sc->esc_k0 - 1) {
sc->esc_n0--;
sc->esc_n1++;
} else {
/* Unexpected geometry; fall back to a full resplit */
break;
}
/* Normalize once every set is the smaller size */
if (sc->esc_n0 == 0) {
sc->esc_k0 = sc->esc_k1;
sc->esc_n0 = sc->esc_n1;
sc->esc_k1 = 0;
sc->esc_n1 = 0;
}
stripe_count--;
}
if (stripe_count > nfound) {
/* Resplit from scratch around the OST count we actually have */
stripe_count = nfound;
ec_split_stripes(stripe_count, suggested, sc);
}
if (stripe_count != orig)
CDEBUG(D_OTHER,
"EC data geometry shrunk %u -> %u stripes (n0=%d k0=%d n1=%d k1=%d) to fit %u OSTs\n",
orig, stripe_count, sc->esc_n0, sc->esc_k0,
sc->esc_n1, sc->esc_k1, nfound);
data_comp->llc_stripe_count = stripe_count;
return stripe_count;
}
/**
* lod_ec_add_data() - Add additional data OSTs to data_comp.
* @env: execution environment for this thread
* @lo: LOD object
* @th: transaction handle
* @nfound: current number of found/selected OSTs in data_comp
* @data_comp: the current data comp.
* @stripe: striping created
* @ost_indices: ost indices of striping created
*
* This function is used to fill the remaining OSTs for an ec protected
* data component, data_comp, by reusing OSTs
*/
static __u32 lod_ec_add_data(const struct lu_env *env, struct lod_object *lo,
struct thandle *th,
__u32 nfound,
struct lod_layout_component *data_comp,
struct dt_object **stripe,
__u32 *ost_indices)
{
int pos = 0;
struct lod_device *lod = lu2lod_dev(lo->ldo_obj.do_lu.lo_dev);
struct dt_object *o;
if (nfound == 0)
return nfound;
while (nfound < data_comp->llc_stripe_count) {
o = lod_qos_declare_object_on(env, lod, ost_indices[pos],
true, th);
if (IS_ERR(o)) {
CDEBUG(D_OTHER, "can't declare object on #%u: %d\n",
ost_indices[pos], (int) PTR_ERR(o));
return nfound;
}
stripe[nfound] = o;
ost_indices[nfound] = ost_indices[pos];
nfound++;
pos++;
}
return nfound;
}
/**
* lod_ost_alloc_rr() - Allocate a striping using round-robin algorithm.
* @env: execution environment for this thread
* @lo: LOD object
* @stripe: striping created [out]
* @ost_indices: ost indices of striping created [out]
* @flags: allocation flags (0 or LOD_USES_DEFAULT_STRIPE)
* @th: transaction handle
* @comp_idx: index of ldo_comp_entries
* @reserve: space to reserve on the target device
* @sc: split_stripes
* @data_comp: data component if this is a parity. NULL otherwise
*
* Allocates a new striping using round-robin algorithm. The function refreshes
* all the internal structures (statfs cache, array of available OSTs sorted
* with regard to OSS, etc). The number of stripes required is taken from the
* object (must be prepared by the caller), but can change if the flag
* LOD_USES_DEFAULT_STRIPE is supplied. The caller should ensure nobody else
* is trying to create a striping on the object in parallel. All the internal
* structures (like pools, etc) are protected and no additional locking is
* required. The function succeeds even if a single stripe is allocated. To save
* time we give priority to targets which already have objects precreated.
* Full OSTs are skipped (see lod_qos_dev_is_full() for the details).
*
* Return:
* * %0 on success
* * %-ENOSPC if not enough OSTs are found
* * %negative negated errno for other failures
*/
static int lod_ost_alloc_rr(const struct lu_env *env, struct lod_object *lo,
struct dt_object **stripe, __u32 *ost_indices,
enum lod_uses_hint flags, struct thandle *th,
int comp_idx, __u64 reserve,
struct ec_split_comp *sc,
struct lod_layout_component *data_comp)
{
struct lod_layout_component *lod_comp;
struct lod_device *m = lu2lod_dev(lo->ldo_obj.do_lu.lo_dev);
struct lod_pool_desc *pool = NULL;
struct lu_tgt_pool *osts;
struct lu_qos_rr *lqr;
unsigned int i, array_idx;
__u32 stripe_idx = 0;
__u32 stripe_count, stripe_count_min, ost_idx;
int rc, speed = 0, ost_connecting = 0;
int idx, stripes_per_ost = 1;
bool overstriped = false;
bool want_overstriping;
ENTRY;
LASSERT(lo->ldo_comp_cnt > comp_idx && lo->ldo_comp_entries != NULL);
lod_comp = &lo->ldo_comp_entries[comp_idx];
stripe_count = lod_comp->llc_stripe_count;
stripe_count_min = lod_stripe_count_min(lod_comp, flags, sc, false);
want_overstriping = lod_comp->llc_pattern & LOV_PATTERN_OVERSTRIPING;
if (lod_comp->llc_pool != NULL)
pool = lod_find_pool(m, lod_comp->llc_pool);
if (pool != NULL) {
down_read(&pool_tgt_rw_sem(pool));
osts = &(pool->pool_obds);
lqr = &(pool->pool_rr);
} else {
osts = &m->lod_ost_descs.ltd_tgt_pool;
lqr = &(m->lod_ost_descs.ltd_qos.lq_rr);
}
rc = lod_qos_calc_rr(m, &m->lod_ost_descs, osts, lqr);
if (rc)
GOTO(out_pool, rc);
rc = lod_qos_tgt_in_use_clear(env, stripe_count);
if (rc)
GOTO(out_pool, rc);
down_read(&m->lod_ost_descs.ltd_qos.lq_rw_sem);
spin_lock(&lqr->lqr_alloc);
if (--lqr->lqr_start_count <= 0) {
atomic_set(&lqr->lqr_start_idx,
get_random_u32_below(osts->op_count));
lqr->lqr_start_count =
(LOV_CREATE_RESEED_MIN / max(osts->op_count, 1U) +
LOV_CREATE_RESEED_MULT) * max(osts->op_count, 1U);
} else if (atomic_read(&lqr->lqr_start_idx) >= osts->op_count) {
/* If we have allocated from all of the tgts, slowly
* precess the next start OST if the tgt/stripe count
* difference isn't already doing this for us.
*/
atomic_sub(osts->op_count, &lqr->lqr_start_idx);
if (stripe_count > 1 && (osts->op_count % stripe_count) != 1)
++lqr->lqr_offset_idx;
}
spin_unlock(&lqr->lqr_alloc);
if (lod_comp->llc_pattern & LOV_PATTERN_OVERSTRIPING)
stripes_per_ost =
(lod_comp->llc_stripe_count - 1) / osts->op_count + 1;
repeat_find:
CDEBUG(D_OTHER, "pool '%s' want %d start_idx %d start_count %d offset %d active %d count %d\n",
lod_comp->llc_pool ? lod_comp->llc_pool : "",
stripe_count, atomic_read(&lqr->lqr_start_idx),
lqr->lqr_start_count, lqr->lqr_offset_idx, osts->op_count,
osts->op_count);
for (i = 0, idx = 0; i < osts->op_count * stripes_per_ost &&
stripe_idx < stripe_count; i++) {
if (likely(speed < 2) || i == 0) {
idx = atomic_inc_return(&lqr->lqr_start_idx) +
lqr->lqr_offset_idx;
} else {
/*
* For last speed, use OSTs one by one
*/
idx++;
}
array_idx = idx % osts->op_count;
ost_idx = lqr->lqr_pool.op_array[array_idx];
CDEBUG(D_OTHER, "#%d strt %d act %d strp %d ary %d idx %d\n",
i, idx, /* XXX: active*/ 0,
stripe_idx, array_idx, ost_idx);
if ((ost_idx == LOV_QOS_EMPTY) ||
!test_bit(ost_idx, m->lod_ost_bitmap))
continue;
/* Fail Check before osc_precreate() is called
so we can only 'fail' single OSC. */
if (CFS_FAIL_CHECK(OBD_FAIL_MDS_OSC_PRECREATE) && ost_idx == 0)
continue;
if (CFS_FAIL_PRECHECK(OBD_FAIL_MDS_LOD_CREATE_PAUSE)) {
clear_bit(LQ_SAME_SPACE,
&m->lod_ost_descs.ltd_qos.lq_flags);
CFS_FAIL_TIMEOUT(OBD_FAIL_MDS_LOD_CREATE_PAUSE,
cfs_fail_val);
}
rc = lod_check_and_reserve_ost(env, lo, lod_comp, ost_idx,
speed, &stripe_idx, stripe,
ost_indices, th, &overstriped,
reserve);
if (rc != 0 && OST_TGT(m, ost_idx)->ltd_discon)
ost_connecting = 1;
}
if ((speed < 2) && (stripe_idx < stripe_count_min)) {
/* Try again, allowing slower OSCs */
speed++;
ost_connecting = 0;
goto repeat_find;
}
up_read(&m->lod_ost_descs.ltd_qos.lq_rw_sem);
/* If there are enough OSTs, a component with overstriping requested
* will not actually end up overstriped. The comp should reflect this.
*/
if (!overstriped)
lod_comp->llc_pattern &= ~LOV_PATTERN_OVERSTRIPING;
if (lod_comp->llc_cstripe_count &&
!(lod_comp->llc_flags & LCME_FL_PARITY) &&
!want_overstriping &&
stripe_idx > 0 && stripe_idx < stripe_count &&
4 * stripe_idx >= 3 * stripe_count) {
CDEBUG(D_OTHER,
"EC RR: shrink data stripes %u -> %u (>= 75%% of request)\n",
stripe_count, stripe_idx);
stripe_count = stripe_idx;
stripe_count_min = stripe_count;
lod_comp->llc_stripe_count = stripe_count;
ec_split_stripes(lod_comp->llc_stripe_count,
lod_comp->llc_dstripe_count, sc);
}
if (stripe_idx) {
lod_comp->llc_stripe_count = stripe_idx;
/* at least one stripe is allocated */
rc = 0;
} else {
/* nobody provided us with a single object */
if (ost_connecting)
rc = -EINPROGRESS;
else
rc = -ENOSPC;
}
/* EC protected data/parity pairs need special handling if we could not
* get enough unique OSTs. The EC fallback functions use
* llc_stripe_count as their target, so temporarily restore
* the original requested count.
*/
if (lod_comp->llc_cstripe_count &&
stripe_idx < stripe_count) {
lod_comp->llc_stripe_count = stripe_count;
/* We can only reuse data stripes for parity if there is more
* than one raid set.
*/
if ((lod_comp->llc_flags & LCME_FL_PARITY) &&
(ec_raidset_count(sc) > 1))
stripe_idx = lod_ec_add_parities(env, lo, th,
stripe_idx, sc,
data_comp, lod_comp,
stripe, ost_indices);
/*
* EC-protected data component is short of unique OSTs.
* With overstriping (-C) reuse OSTs from earlier raid
* sets. Without it (-c), shrink the EC geometry by
* converting esc_k0 raidsets into esc_k1 until the
* allocated OST count is enough — prefer a smaller
* layout over failing create when OSTs are offline.
*/
if (!(lod_comp->llc_flags & LCME_FL_PARITY) &&
(ec_raidset_count(sc) > 1) &&
stripe_idx >= stripe_count_min) {
if (want_overstriping) {
stripe_idx = lod_ec_add_data(env, lo,
th,
stripe_idx,
lod_comp,
stripe,
ost_indices);
} else {
/*
* Shrink sets llc_stripe_count and sc
* to a geometry that fits stripe_idx
* unique OSTs. stripe_count tracks
* the new target for the equality
* check below.
*/
stripe_count =
lod_ec_shrink_data_geometry(stripe_idx,
lod_comp,
sc);
}
}
if (stripe_idx == stripe_count) {
lod_comp->llc_stripe_count = stripe_idx;
rc = 0;
} else {
/* Still short; remember how many we got */
lod_comp->llc_stripe_count = stripe_idx;
}
}
if (lod_comp->llc_cstripe_count && stripe_idx < stripe_count_min)
rc = -ENOSPC;
out_pool:
if (pool != NULL) {
up_read(&pool_tgt_rw_sem(pool));
/* put back ref got by lod_find_pool() */
lod_pool_putref(pool);
}
RETURN(rc);
}
static int
lod_qos_mdt_in_use_init(const struct lu_env *env,
const struct lu_tgt_descs *ltd,
u32 stripe_idx, u32 stripe_count,
const struct lu_tgt_pool *pool,
struct dt_object **stripes)
{
u32 mdt_idx;
struct lu_tgt_desc *mdt;
int i, j;
int rc;
rc = lod_qos_tgt_in_use_clear(env, stripe_count);
if (rc)
return rc;
/* if stripe_idx > 1, we are splitting directory, mark existing stripes
* in_use. Because for either split or creation, stripe 0 is local,
* don't mark it in use.
*/
for (i = 1; i < stripe_idx; i++) {
LASSERT(stripes[i]);
for (j = 0; j < pool->op_count; j++) {
mdt_idx = pool->op_array[j];
if (!test_bit(mdt_idx, ltd->ltd_tgt_bitmap))
continue;
mdt = LTD_TGT(ltd, mdt_idx);
if (&mdt->ltd_tgt->dd_lu_dev ==
stripes[i]->do_lu.lo_dev)
lod_qos_tgt_in_use(env, i, mdt_idx);
}
}
return 0;
}
/**
* lod_mdt_alloc_rr() - Allocate a striping using round-robin algorithm.
* @env: execution environment for this thread
* @lo: LOD object
* @stripes: striping created
* @stripe_idx: starting index of stripe allocation
* @stripe_count: Number of stripe objects needed
*
* Allocates a new striping using round-robin algorithm. The function refreshes
* all the internal structures (statfs cache, array of available remote MDTs
* sorted with regard to MDS, etc). The number of stripes required is taken from
* the object (must be prepared by the caller). The caller should ensure nobody
* else is trying to create a striping on the object in parallel. All the
* internal structures (like pools, etc) are protected and no additional locking
* is required. The function succeeds even if a single stripe is allocated.
*
* Return:
* * %positive stripe objects allocated, including first stripe allocated
* outside the function
* * %-ENOSPC if not enough MDTs are found
* * %negative negated errno for other failures
*/
int lod_mdt_alloc_rr(const struct lu_env *env, struct lod_object *lo,
struct dt_object **stripes, u32 stripe_idx,
u32 stripe_count)
{
struct lod_device *lod = lu2lod_dev(lo->ldo_obj.do_lu.lo_dev);
struct lu_tgt_descs *ltd = &lod->lod_mdt_descs;
struct lu_tgt_pool *pool;
struct lu_qos_rr *lqr;
struct lu_object_conf conf = { .loc_flags = LOC_F_NEW };
struct lu_fid fid = { 0 };
struct dt_object *dto;
unsigned int pool_idx;
unsigned int i;
u32 saved_idx = stripe_idx;
int stripes_per_mdt = 1;
u32 mdt_idx;
bool use_degraded = false;
bool overstriped = false;
int tgt_connecting = 0;
int rc;
ENTRY;
pool = <d->ltd_tgt_pool;
lqr = <d->ltd_qos.lq_rr;
rc = lod_qos_calc_rr(lod, ltd, pool, lqr);
if (rc)
RETURN(rc);
overstriped = lo->ldo_dir_hash_type & LMV_HASH_FLAG_OVERSTRIPED;
if (stripe_count > lod->lod_remote_mdt_count + 1 && !overstriped)
RETURN(-E2BIG);
if (lo->ldo_dir_hash_type & LMV_HASH_FLAG_OVERSTRIPED)
stripes_per_mdt = stripe_count / (pool->op_count + 1);
rc = lod_qos_mdt_in_use_init(env, ltd, stripe_idx, stripe_count, pool,
stripes);
if (rc)
RETURN(rc);
down_read(<d->ltd_qos.lq_rw_sem);
spin_lock(&lqr->lqr_alloc);
if (--lqr->lqr_start_count <= 0) {
atomic_set(&lqr->lqr_start_idx,
get_random_u32_below(pool->op_count));
lqr->lqr_start_count =
(LOV_CREATE_RESEED_MIN / max(pool->op_count, 1U) +
LOV_CREATE_RESEED_MULT) * max(pool->op_count, 1U);
} else if (atomic_read(&lqr->lqr_start_idx) >= pool->op_count) {
/* If we have allocated from all of the tgts, slowly
* precess the next start if the tgt/stripe count isn't
* already doing this for us.
*/
atomic_sub(pool->op_count, &lqr->lqr_start_idx);
if (stripe_count - 1 > 1 &&
(pool->op_count % (stripe_count - 1)) != 1)
++lqr->lqr_offset_idx;
}
spin_unlock(&lqr->lqr_alloc);
repeat_find:
CDEBUG(D_OTHER,
"want=%d start_idx=%d start_count=%d offset=%d active=%d count=%d\n",
stripe_count - 1, atomic_read(&lqr->lqr_start_idx),
lqr->lqr_start_count, lqr->lqr_offset_idx,
/* if we're overstriped, the local MDT is available and is
* included in the count
*/
pool->op_count + overstriped,
lqr->lqr_pool.op_count + overstriped);
for (i = 0; i < (pool->op_count + overstriped) * stripes_per_mdt &&
stripe_idx < stripe_count; i++) {
struct lu_tgt_desc *mdt = NULL;
struct dt_device *mdt_tgt;
bool local_alloc = false;
int idx;
idx = atomic_inc_return(&lqr->lqr_start_idx);
pool_idx = (idx + lqr->lqr_offset_idx) %
(pool->op_count + overstriped);
/* in the overstriped case, we must be able to allocate a stripe
* to the local MDT, ie, the one doing the allocation
*/
if (pool_idx == pool->op_count) {
LASSERT(overstriped);
/* because there is already a stripe on the local MDT,
* do not allocate from the local MDT until we've
* allocated at least as many stripes as we have MDTs
*/
if (stripe_idx < (pool->op_count + 1)) {
CDEBUG(D_OTHER,
"Skipping local alloc, not enough stripes yet\n");
continue;
}
CDEBUG(D_OTHER, "Attempting to allocate locally\n");
local_alloc = true;
mdt_tgt = lod->lod_child;
rc = lodname2mdt_index(lod2obd(lod)->obd_name,
&mdt_idx);
/* this parsing can't fail here because we're working
* with a known-good MDT
*/
LASSERT(!rc);
} else {
mdt_idx = lqr->lqr_pool.op_array[pool_idx];
mdt = LTD_TGT(ltd, mdt_idx);
mdt_tgt = mdt->ltd_tgt;
}
CDEBUG(D_OTHER, "#%d strt %d act %d strp %d ary %d idx %d\n",
i, idx, /* XXX: active*/ 0,
stripe_idx, pool_idx, mdt_idx);
if (!local_alloc && (mdt_idx == LOV_QOS_EMPTY ||
!test_bit(mdt_idx, ltd->ltd_tgt_bitmap))) {
CDEBUG(D_OTHER, "mdt_idx not found %d\n", mdt_idx);
continue;
}
/* do not put >1 objects on one MDT, except for overstriping */
if (!local_alloc) {
if (lo->ldo_dir_hash_type & LMV_HASH_FLAG_OVERSTRIPED) {
CDEBUG(D_OTHER, "overstriped\n");
} else if (lod_qos_is_tgt_used(env, mdt_idx,
stripe_idx)) {
CDEBUG(D_OTHER, "#%d: already used\n", mdt_idx);
continue;
}
}
/* we know the local MDT is usable */
if (!local_alloc) {
if (mdt->ltd_discon) {
tgt_connecting = 1;
CDEBUG(D_OTHER, "#%d: unusable\n", mdt_idx);
continue;
}
if (lod_statfs_check(ltd, mdt))
continue;
if (mdt->ltd_statfs.os_state & OS_STATFS_NOCREATE)
continue;
}
/* try to use another OSP if this one is degraded */
if (!local_alloc && !use_degraded &&
mdt->ltd_statfs.os_state & OS_STATFS_DEGRADED) {
CDEBUG(D_OTHER, "#%d: degraded\n", mdt_idx);
continue;
}
rc = dt_fid_alloc(env, mdt_tgt, &fid, NULL, NULL);
if (rc < 0) {
CDEBUG(D_OTHER, "#%d: alloc FID failed: %dl\n", mdt_idx, rc);
continue;
}
dto = dt_locate_at(env, mdt_tgt, &fid,
lo->ldo_obj.do_lu.lo_dev->ld_site->ls_top_dev,
&conf);
if (IS_ERR(dto)) {
CDEBUG(D_OTHER, "can't alloc stripe on #%u: %d\n",
mdt_idx, (int) PTR_ERR(dto));
if (!local_alloc && mdt->ltd_discon)
tgt_connecting = 1;
continue;
}
lod_qos_tgt_in_use(env, stripe_idx, mdt_idx);
stripes[stripe_idx++] = dto;
}
if (!use_degraded && stripe_idx < stripe_count) {
/* Try again, allowing slower MDTs */
use_degraded = true;
tgt_connecting = 0;
goto repeat_find;
}
up_read(<d->ltd_qos.lq_rw_sem);
if (stripe_idx > saved_idx) {
/* If there are enough MDTs, we will not actually do
* overstriping, and the hash flags should reflect this.
*/
if (!overstriped)
lo->ldo_dir_hash_type &= ~LMV_HASH_FLAG_OVERSTRIPED;
/* at least one stripe is allocated */
RETURN(stripe_idx);
}
/* nobody provided us with a single object */
if (tgt_connecting)
RETURN(-EINPROGRESS);
RETURN(-ENOSPC);
}
/**
* lod_alloc_ost_list() - Allocate a specific striping layout on a user defined
* set of OSTs.
* @env: execution environment for this thread
* @lo: LOD object
* @stripe: striping created [out]
* @ost_indices: ost indices of striping created [out]
* @th: transaction handle
* @comp_idx: index of ldo_comp_entries
* @reserve: space to reserve on the target device
*
* Allocates new striping using the OST index range provided by the data from
* the lmm_objects contained in the lov_user_md passed to this method. Full
* OSTs are not considered. The exact order of OSTs requested by the user
* is respected as much as possible depending on OST status. The number of
* stripes needed and stripe offset are taken from the object. If that number
* can not be met, then the function returns a failure and then it's the
* caller's responsibility to release the stripes allocated. All the internal
* structures are protected, but no concurrent allocation is allowed on the
* same objects.
*
* Return:
* * %0 on success
* * %-ENODEV OST index does not exist on file system
* * %-EINVAL requested OST index is invalid
* * %negative negated errno on error
*/
static int lod_alloc_ost_list(const struct lu_env *env, struct lod_object *lo,
struct dt_object **stripe, __u32 *ost_indices,
struct thandle *th, int comp_idx, __u64 reserve)
{
struct lod_layout_component *lod_comp;
struct lod_device *m = lu2lod_dev(lo->ldo_obj.do_lu.lo_dev);
struct dt_object *o;
unsigned int array_idx = 0;
int stripe_count = 0;
int i;
int rc = -EINVAL;
ENTRY;
/* for specific OSTs layout */
LASSERT(lo->ldo_comp_cnt > comp_idx && lo->ldo_comp_entries != NULL);
lod_comp = &lo->ldo_comp_entries[comp_idx];
LASSERT(lod_comp->llc_ostlist.op_array);
LASSERT(lod_comp->llc_ostlist.op_count);
rc = lod_qos_tgt_in_use_clear(env, lod_comp->llc_stripe_count);
if (rc < 0)
RETURN(rc);
if (lod_comp->llc_stripe_offset == LOV_OFFSET_DEFAULT)
lod_comp->llc_stripe_offset =
lod_comp->llc_ostlist.op_array[0];
for (i = 0; i < lod_comp->llc_stripe_count; i++) {
if (lod_comp->llc_ostlist.op_array[i] ==
lod_comp->llc_stripe_offset) {
array_idx = i;
break;
}
}
if (i == lod_comp->llc_stripe_count) {
CDEBUG(D_OTHER,
"%s: start index %d not in the specified list of OSTs\n",
lod2obd(m)->obd_name, lod_comp->llc_stripe_offset);
RETURN(-EINVAL);
}
for (i = 0; i < lod_comp->llc_stripe_count;
i++, array_idx = (array_idx + 1) % lod_comp->llc_stripe_count) {
__u32 ost_idx = lod_comp->llc_ostlist.op_array[array_idx];
if (!test_bit(ost_idx, m->lod_ost_bitmap)) {
rc = -EINVAL;
break;
}
/* do not put >1 objects on a single OST, except for
* overstriping
*/
if (lod_qos_is_tgt_used(env, ost_idx, stripe_count) &&
!(lod_comp->llc_pattern & LOV_PATTERN_OVERSTRIPING)) {
rc = -EINVAL;
break;
}
rc = lod_statfs_and_check(env, m, &m->lod_ost_descs,
LTD_TGT(&m->lod_ost_descs, ost_idx),
reserve);
if (rc < 0) /* this OSP doesn't feel well */
break;
o = lod_qos_declare_object_on(env, m, ost_idx, true, th);
if (IS_ERR(o)) {
rc = PTR_ERR(o);
CDEBUG(D_OTHER,
"%s: can't declare new object on #%u: %d\n",
lod2obd(m)->obd_name, ost_idx, rc);
break;
}
/*
* We've successfully declared (reserved) an object
*/
lod_qos_tgt_in_use(env, stripe_count, ost_idx);
stripe[stripe_count] = o;
ost_indices[stripe_count] = ost_idx;
stripe_count++;
}
RETURN(rc);
}
/**
* lod_ost_alloc_specific() - Allocate a striping on a predefined set of OSTs.
* @env: execution environment for this thread
* @lo: LOD object
* @stripe: striping created [out]
* @ost_indices: ost indices of striping created [out]
* @flags: not used
* @th: transaction handle
* @comp_idx: index of ldo_comp_entries
* @reserve: space to reserve on the target device
*
* Allocates new layout starting from OST index in lo->ldo_stripe_offset.
* Full OSTs are not considered. The exact order of OSTs is not important and
* varies depending on OST status. The allocation procedure prefers the targets
* with precreated objects ready. The number of stripes needed and stripe
* offset are taken from the object. If that number cannot be met, then the
* function returns an error and then it's the caller's responsibility to
* release the stripes allocated. All the internal structures are protected,
* but no concurrent allocation is allowed on the same objects.
*
* Return:
* * %0 on success
* * %-ENOSPC if no OST objects are available at all
* * %-EFBIG if not enough OST objects are found
* * %-EINVAL requested offset is invalid
* * %negative errno on failure
*/
static int lod_ost_alloc_specific(const struct lu_env *env,
struct lod_object *lo,
struct dt_object **stripe, __u32 *ost_indices,
enum lod_uses_hint flags, struct thandle *th,
int comp_idx, __u64 reserve)
{
struct lod_layout_component *lod_comp;
struct lod_device *m = lu2lod_dev(lo->ldo_obj.do_lu.lo_dev);
struct dt_object *o;
struct lu_tgt_desc *tgt;
__u32 ost_idx;
unsigned int i, array_idx, ost_count;
int rc, stripe_num = 0;
int speed = 0;
struct lod_pool_desc *pool = NULL;
struct lu_tgt_pool *osts;
int stripes_per_ost = 1;
bool overstriped = false;
ENTRY;
LASSERT(lo->ldo_comp_cnt > comp_idx && lo->ldo_comp_entries != NULL);
lod_comp = &lo->ldo_comp_entries[comp_idx];
rc = lod_qos_tgt_in_use_clear(env, lod_comp->llc_stripe_count);
if (rc)
GOTO(out, rc);
if (lod_comp->llc_pool != NULL)
pool = lod_find_pool(m, lod_comp->llc_pool);
if (pool != NULL) {
down_read(&pool_tgt_rw_sem(pool));
osts = &(pool->pool_obds);
} else {
osts = &m->lod_ost_descs.ltd_tgt_pool;
}
ost_count = osts->op_count;
repeat_find:
/* search loi_ost_idx in ost array */
array_idx = 0;
for (i = 0; i < ost_count; i++) {
if (osts->op_array[i] == lod_comp->llc_stripe_offset) {
array_idx = i;
break;
}
}
if (i == ost_count) {
/*
* For uninitialized PFL components, the stripe offset may be
* invalid (e.g., pointing to a deactivated OST). In this case,
* fall back to QoS allocation instead of failing. This can
* happen during migration when old layouts have deprecated
* stripe offsets stored in lmm_layout_gen.
*/
if (!lod_comp_inited(lod_comp)) {
CDEBUG(D_LAYOUT,
"Uninitialized component %d has invalid start index %d, using QoS allocation\n",
comp_idx, lod_comp->llc_stripe_offset);
lod_comp->llc_stripe_offset = LOV_OFFSET_DEFAULT;
GOTO(out, rc = -EAGAIN);
}
CERROR("Start index %d not found in pool '%s'\n",
lod_comp->llc_stripe_offset,
lod_comp->llc_pool ? lod_comp->llc_pool : "");
GOTO(out, rc = -EINVAL);
}
if (lod_comp->llc_pattern & LOV_PATTERN_OVERSTRIPING)
stripes_per_ost =
(lod_comp->llc_stripe_count - 1)/ost_count + 1;
/* user specifies bigger stripe count than available ost count */
if (lod_comp->llc_stripe_count > ost_count * stripes_per_ost)
lod_comp->llc_stripe_count = ost_count * stripes_per_ost;
for (i = 0; i < ost_count * stripes_per_ost;
i++, array_idx = (array_idx + 1) % ost_count) {
ost_idx = osts->op_array[array_idx];
if (!test_bit(ost_idx, m->lod_ost_bitmap))
continue;
/* Fail Check before osc_precreate() is called
so we can only 'fail' single OSC. */
if (CFS_FAIL_CHECK(OBD_FAIL_MDS_OSC_PRECREATE) && ost_idx == 0)
continue;
/*
* do not put >1 objects on a single OST, except for
* overstriping, where it is intended
*/
if (lod_qos_is_tgt_used(env, ost_idx, stripe_num)) {
if (lod_comp->llc_pattern & LOV_PATTERN_OVERSTRIPING)
overstriped = true;
else
continue;
}
/*
* try not allocate on the OST used by other component
*/
if (speed == 0 && i != 0 &&
lod_comp_is_ost_used(env, lo, ost_idx))
continue;
tgt = LTD_TGT(&m->lod_ost_descs, ost_idx);
/* Drop slow OSCs if we can, but not for requested start idx.
*
* This means "if OSC is slow and it is not the requested
* start OST, then it can be skipped, otherwise skip it only
* if it is inactive/recovering/out-of-space." */
rc = lod_statfs_and_check(env, m, &m->lod_ost_descs,
tgt, reserve);
if (rc) {
/* this OSP doesn't feel well */
continue;
}
/*
* We expect number of precreated objects at the first
* iteration. Skip OSPs with no objects ready. Don't apply
* this logic to OST specified with stripe_offset.
*/
if (i && !tgt->ltd_statfs.os_fprecreated && !speed)
continue;
o = lod_qos_declare_object_on(env, m, ost_idx, true, th);
if (IS_ERR(o)) {
CDEBUG(D_OTHER, "can't declare new object on #%u: %d\n",
ost_idx, (int) PTR_ERR(o));
continue;
}
/*
* We've successfully declared (reserved) an object
*/
lod_qos_tgt_in_use(env, stripe_num, ost_idx);
stripe[stripe_num] = o;
ost_indices[stripe_num] = ost_idx;
stripe_num++;
/* We have enough stripes */
if (stripe_num == lod_comp->llc_stripe_count)
GOTO(out, rc = 0);
}
if (speed < 2) {
/* Try again, allowing slower OSCs */
speed++;
goto repeat_find;
}
/* If we were passed specific striping params, then a failure to
* meet those requirements is an error, since we can't reallocate
* that memory (it might be part of a larger array or something).
*/
CERROR("can't lstripe objid "DFID": have %d want %u\n",
PFID(lu_object_fid(lod2lu_obj(lo))), stripe_num,
lod_comp->llc_stripe_count);
rc = stripe_num == 0 ? -ENOSPC : -EFBIG;
/* If there are enough OSTs, a component with overstriping requessted
* will not actually end up overstriped. The comp should reflect this.
*/
if (rc == 0 && !overstriped)
lod_comp->llc_pattern &= ~LOV_PATTERN_OVERSTRIPING;
out:
if (pool != NULL) {
up_read(&pool_tgt_rw_sem(pool));
/* put back ref got by lod_find_pool() */
lod_pool_putref(pool);
}
RETURN(rc);
}
struct semaphore_timer {
struct timer_list timer;
struct task_struct *task;
};
static void process_semaphore_timer(struct timer_list *t)
{
struct semaphore_timer *timeout = cfs_from_timer(timeout, t, timer);
send_sig(SIGKILL, timeout->task, 1);
}
/* Whether QoS data in pool is up-to-date and balanced. */
static bool pool_qos_is_usable(struct lod_pool_desc *pool)
{
time64_t now;
now = ktime_get_real_seconds();
if (pool->pool_same_space && now < pool->pool_same_space_expire)
return false;
return true;
}
/**
* lod_pool_qos_penalties_calc() - Calculate penalties per-ost in a pool
* @lod: lod_device
* @pool: pool_desc
*
* The algorithm is similar to ltd_qos_penalties_calc(), but much simpler,
* just considering the space of each OST in this pool.
*
* Return:
* * %0 on success
* * %-EAGAIN the number of OSTs isn't enough or all tgt spaces are almost the
* same
*/
static int lod_pool_qos_penalties_calc(struct lod_device *lod,
struct lod_pool_desc *pool)
{
struct lu_tgt_descs *ltd = &lod->lod_ost_descs;
struct lu_qos *qos = <d->ltd_qos;
struct lov_desc *desc = <d->ltd_lov_desc;
struct lu_tgt_pool *osts = &pool->pool_obds;
struct lod_tgt_desc *ost;
__u64 ba_max, ba_min, ba;
__u32 num_active;
int prio_wide;
time64_t now, age;
int i, rc;
ENTRY;
num_active = osts->op_count - 1;
if (num_active < 1)
GOTO(out, rc = -EAGAIN);
prio_wide = 256 - qos->lq_prio_free;
ba_min = (__u64)(-1);
ba_max = 0;
now = ktime_get_real_seconds();
/* Calculate penalty per OST */
for (i = 0; i < osts->op_count; i++) {
if (!test_bit(osts->op_array[i], lod->lod_ost_bitmap))
continue;
ost = OST_TGT(lod, osts->op_array[i]);
if (!ost->ltd_active)
continue;
ba = tgt_statfs_bavail(ost) >> 8;
if (!ba)
continue;
ba_min = min(ba, ba_min);
ba_max = max(ba, ba_max);
ost->ltd_qos.ltq_svr->lsq_bavail += ba;
/*
* per-ost penalty is
* prio * bavail / (num_tgt - 1) / prio_max / 2
*/
ost->ltd_qos.ltq_penalty_per_obj = prio_wide * ba >> 9;
do_div(ost->ltd_qos.ltq_penalty_per_obj, num_active);
age = (now - ost->ltd_qos.ltq_used) >> 3;
if (age > 32 * desc->ld_qos_maxage)
ost->ltd_qos.ltq_penalty = 0;
else if (age > desc->ld_qos_maxage)
/* Decay ost penalty. */
ost->ltd_qos.ltq_penalty >>= age / desc->ld_qos_maxage;
}
/*
* If each ost has almost same free space, do rr allocation for better
* creation performance
*/
if ((ba_max * (256 - qos->lq_threshold_rr)) >> 8 < ba_min) {
pool->pool_same_space = true;
pool->pool_same_space_expire = now + desc->ld_qos_maxage;
} else {
pool->pool_same_space = false;
}
rc = 0;
out:
if (!rc && pool->pool_same_space)
rc = -EAGAIN;
RETURN(rc);
}
/**
* lod_ost_alloc_qos() - Allocate a striping using an algorithm with weights.
* @env: execution environment for this thread
* @lo: LOD object
* @stripe: striping created
* @ost_indices: ost indices of striping created
* @flags: 0 or LOD_USES_DEFAULT_STRIPE
* @th: transaction handle
* @comp_idx: index of ldo_comp_entries
* @reserve: space to reserve on the target device
* @sc: split_stripes
* @data_comp: data component if this is a parity. NULL otherwise/
*
* The function allocates OST objects to create a striping. The algorithm
* used is based on weights (currently only using the free space), and it's
* trying to ensure the space is used evenly by OSTs and OSSs. The striping
* configuration (# of stripes, offset, pool) is taken from the object and
* is prepared by the caller.
*
* If LOD_USES_DEFAULT_STRIPE is not passed and prepared configuration can't
* be met due to too few OSTs, then allocation fails. If the flag is passed
* fewer than 3/4 of the requested number of stripes can be allocated, then
* allocation fails.
*
* No concurrent allocation is allowed on the object and this must be ensured
* by the caller. All the internal structures are protected by the function.
*
* The algorithm has two steps: find available OSTs and calculate their
* weights, then select the OSTs with their weights used as the probability.
* An OST with a higher weight is proportionately more likely to be selected
* than one with a lower weight.
*
* Return:
* * %0 on success
* * %-EAGAIN not enough OSTs are found for specified stripe count
* * %-EINVAL requested OST index is invalid
* * %negative errno on failure
*/
static int lod_ost_alloc_qos(const struct lu_env *env, struct lod_object *lo,
struct dt_object **stripe, __u32 *ost_indices,
enum lod_uses_hint flags, struct thandle *th,
int comp_idx, __u64 reserve,
struct ec_split_comp *sc,
struct lod_layout_component *data_comp)
{
struct lod_layout_component *lod_comp;
struct lod_device *lod = lu2lod_dev(lo->ldo_obj.do_lu.lo_dev);
struct lod_avoid_guide *lag = &lod_env_info(env)->lti_avoid;
struct lod_tgt_desc *ost;
struct dt_object *o;
__u64 total_weight = 0;
struct lod_pool_desc *pool = NULL;
struct lu_tgt_pool *osts;
unsigned int i;
__u32 nfound, good_osts, stripe_count, stripe_count_min;
bool overstriped = false;
int stripes_per_ost = 1;
bool slow = false;
int rc = 0;
ENTRY;
/* Totally skip qos part when qos_threshold_rr=100% */
if (lod->lod_ost_descs.ltd_qos.lq_threshold_rr == QOS_THRESHOLD_MAX)
return -EAGAIN;
LASSERT(lo->ldo_comp_cnt > comp_idx && lo->ldo_comp_entries != NULL);
lod_comp = &lo->ldo_comp_entries[comp_idx];
stripe_count_min = lod_stripe_count_min(lod_comp, flags, sc, true);
stripe_count = lod_comp->llc_stripe_count;
/* Orphaned parity component? */
if ((lod_comp->llc_flags & LCME_FL_PARITY) && data_comp == NULL)
RETURN(-EINVAL);
/* For parities we allow stripe_count_min==0 as we will reuse OSTs */
if (stripe_count_min < 1 && (!(lod_comp->llc_flags & LCME_FL_PARITY)))
RETURN(-EINVAL);
if (lod_comp->llc_pool != NULL)
pool = lod_find_pool(lod, lod_comp->llc_pool);
/* Detect -EAGAIN early, before expensive qos write lock is taken. */
if (pool) {
down_read(&pool_tgt_rw_sem(pool));
if (!pool_qos_is_usable(pool))
GOTO(out_nolock, rc = -EAGAIN);
osts = &(pool->pool_obds);
} else {
if (!ltd_qos_is_usable(&lod->lod_ost_descs))
GOTO(out_nolock, rc = -EAGAIN);
osts = &lod->lod_ost_descs.ltd_tgt_pool;
}
if (lod_comp->llc_pattern & LOV_PATTERN_OVERSTRIPING)
stripes_per_ost =
(lod_comp->llc_stripe_count - 1)/osts->op_count + 1;
if (!down_write_trylock(&lod->lod_ost_descs.ltd_qos.lq_rw_sem)) {
struct semaphore_timer timer;
/*
* The timer wakes us from down_write_killable() with
* send_sig(SIGKILL). Set the SIGKILL disposition to
* SIG_KTHREAD: the kthread default SIG_IGN would discard
* the signal, and SIG_DFL would make sig_fatal() latch
* SIGNAL_GROUP_EXIT on this long-lived service thread at
* the first wakeup, after which kernels >= v6.1 silently
* drop all further signals to this task, leaving later
* waiters stuck for as long as lq_rw_sem is held.
*/
allow_signal(SIGKILL);
timer.task = current;
cfs_timer_setup(&timer.timer, process_semaphore_timer, 0, 0);
mod_timer(&timer.timer, jiffies + cfs_time_seconds(2));
/* Do actual allocation, use write lock here. */
rc = down_write_killable(&lod->lod_ost_descs.ltd_qos.lq_rw_sem);
timer_delete_sync(&timer.timer);
disallow_signal(SIGKILL);
if (rc) {
flush_signals(current);
CDEBUG(D_OTHER, "%s: wakeup semaphore on timeout rc = %d\n",
lod2obd(lod)->obd_name, rc);
GOTO(out_nolock, rc = -EAGAIN);
}
}
/*
* Check again, while we were sleeping on @lq_rw_sem things could
* change.
*/
if (pool) {
if (!pool_qos_is_usable(pool))
GOTO(out, rc = -EAGAIN);
rc = lod_pool_qos_penalties_calc(lod, pool);
} else {
if (!ltd_qos_is_usable(&lod->lod_ost_descs))
GOTO(out, rc = -EAGAIN);
rc = ltd_qos_penalties_calc(&lod->lod_ost_descs);
}
if (rc)
GOTO(out, rc);
rc = lod_qos_tgt_in_use_clear(env, lod_comp->llc_stripe_count);
if (rc)
GOTO(out, rc);
good_osts = 0;
/* Find all the OSTs that are valid stripe candidates */
for (i = 0; i < osts->op_count; i++) {
if (!test_bit(osts->op_array[i], lod->lod_ost_bitmap))
continue;
ost = OST_TGT(lod, osts->op_array[i]);
ost->ltd_qos.ltq_usable = 0;
rc = lod_statfs_and_check(env, lod, &lod->lod_ost_descs,
ost, reserve);
if (rc) {
/* this OSP doesn't feel well */
continue;
}
if (ost->ltd_statfs.os_state & OS_STATFS_DEGRADED)
continue;
/* Fail Check before osc_precreate() is called
* so we can only 'fail' single OSC.
*/
if (CFS_FAIL_CHECK(OBD_FAIL_MDS_OSC_PRECREATE) &&
osts->op_array[i] == 0)
continue;
if (lod_comp->llc_cstripe_count &&
lod_should_avoid_ost(lo, lag, osts->op_array[i]))
continue;
ost->ltd_qos.ltq_usable = 1;
lu_tgt_qos_weight_calc(ost, false);
total_weight += ost->ltd_qos.ltq_weight;
good_osts++;
}
/* For EC, if we do not have enough OSTs for the full requested
* number of stripes we can still permit the allocation as
* long as we can allocate at least 75% of the requested number
* of data stripes and one single stripe per parity.
* I.e.
* good >= 0.75 * stripe_count + cstripe_count
* =>
* 4 x good >= 3 * stripe_count + 4 * cstripe_count
*/
if (lod_comp->llc_cstripe_count &&
!(lod_comp->llc_flags & LCME_FL_PARITY) &&
!(lod_comp->llc_pattern & LOV_PATTERN_OVERSTRIPING) &&
good_osts < (lod_comp->llc_stripe_count +
lod_comp->llc_cstripe_count) &&
4 * good_osts >= (3 * lod_comp->llc_stripe_count +
4 * lod_comp->llc_cstripe_count)){
stripe_count = good_osts - lod_comp->llc_cstripe_count;
stripe_count_min = stripe_count;
lod_comp->llc_stripe_count = stripe_count;
/* stripe count changed so we must re-compute the splits */
ec_split_stripes(lod_comp->llc_stripe_count,
lod_comp->llc_dstripe_count,
sc);
}
CDEBUG(D_OTHER, "found %d good osts\n", good_osts);
if (good_osts < stripe_count_min)
GOTO(out, rc = -EAGAIN);
/* If we do not have enough OSTs for the requested stripe count, do not
* put more stripes per OST than requested.
* For ec protected data stripes we ALWAYS allocate the full stripeset.
*/
if (!lod_comp->llc_cstripe_count &&
stripe_count / stripes_per_ost > good_osts)
stripe_count = good_osts * stripes_per_ost;
/* Find enough OSTs with weighted random allocation. */
nfound = 0;
while (nfound < stripe_count) {
u64 rand, cur_weight;
cur_weight = 0;
rc = -ENOSPC;
rand = lu_prandom_u64_max(total_weight);
/* On average, this will hit larger-weighted OSTs more often.
* 0-weight OSTs will always get used last (only when rand=0)
*/
for (i = 0; i < osts->op_count; i++) {
__u32 idx = osts->op_array[i];
struct lod_tgt_desc *ost;
if (lod_should_avoid_ost(lo, lag, idx))
continue;
ost = OST_TGT(lod, idx);
if (!ost->ltd_qos.ltq_usable)
continue;
cur_weight += ost->ltd_qos.ltq_weight;
CDEBUG(D_OTHER,
"stripe_count=%d nfound=%d cur_weight=%llu rand=%llu total_weight=%llu\n",
stripe_count, nfound, cur_weight, rand,
total_weight);
/* TODO: this check is wrong and can cause the qos
* allocator to fail.the problem is that we do not
* compute total_weight and cur_weight the same way
* when OSTs are offline.
* Very annoying since if OSTs are offline this will
* fail, and fallback to rr quite often, making
* testing painful.
*/
if (cur_weight < rand)
continue;
CDEBUG(D_OTHER, "stripe=%d to idx=%d\n", nfound, idx);
/*
* In case of QOS it makes sense to check components
* only for FLR and if current component doesn't
* support overstriping.
*/
if (lo->ldo_mirror_count > 1 &&
!(lod_comp->llc_pattern & LOV_PATTERN_OVERSTRIPING)
&& lod_comp_is_ost_used(env, lo, idx))
continue;
if (lod_qos_is_tgt_used(env, idx, nfound)) {
if (lod_comp->llc_pattern &
LOV_PATTERN_OVERSTRIPING)
overstriped = true;
else
continue;
}
if ((lod_comp->llc_flags & LCME_FL_PARITY) &&
data_comp &&
!lod_ec_check_data_usable(idx, nfound, sc,
data_comp, lod_comp))
continue;
o = lod_qos_declare_object_on(env, lod, idx, slow, th);
if (IS_ERR(o)) {
CDEBUG(D_OTHER,
"can't declare object on #%u:%d\n",
idx, (int) PTR_ERR(o));
continue;
}
lod_avoid_update(lo, lag);
lod_qos_tgt_in_use(env, nfound, idx);
stripe[nfound] = o;
ost_indices[nfound] = idx;
ltd_qos_update(&lod->lod_ost_descs, ost, &total_weight);
nfound++;
rc = 0;
break;
}
if (rc && !slow && nfound < stripe_count) {
/* couldn't allocate using precreated objects
* so try to wait for new precreations */
slow = true;
rc = 0;
}
if (rc) {
/* no OST found on this iteration, give up */
break;
}
}
/* EC protected data/parity pairs need special handling if we could not
* get enough unique OSTs.
*/
if (lod_comp->llc_cstripe_count && nfound < stripe_count &&
nfound >= stripe_count_min) {
if (lod_comp->llc_flags & LCME_FL_PARITY) {
/* If this was a parity allocation we have already
* allocated as many unique OSTs as we could. If there
* are still missing, try to re-use OSTs from the
* corresponding data component.
*/
nfound = lod_ec_add_parities(env, lo, th, nfound, sc,
data_comp, lod_comp,
stripe, ost_indices);
} else if (lod_comp->llc_pattern & LOV_PATTERN_OVERSTRIPING) {
/* Overstriping (-C): reuse OSTs from earlier raid
* sets until the requested stripe count is met.
*/
nfound = lod_ec_add_data(env, lo, th, nfound, lod_comp,
stripe, ost_indices);
} else {
/*
* No overstriping (-c): shrink the EC geometry by
* converting esc_k0 raidsets into esc_k1 until the
* unique OSTs already allocated are enough. Prefer
* a smaller layout over failing create when OSTs
* are offline.
*/
stripe_count =
lod_ec_shrink_data_geometry(nfound,
lod_comp, sc);
}
if (nfound == lod_comp->llc_stripe_count)
rc = 0;
}
/* For EC components we require the full stripe_count to avoid
* incomplete raid sets. For plain components we allow partial
* allocations down to stripe_count_min (e.g., 3/4 of the request
* for LOD_USES_DEFAULT_STRIPE) to preserve QoS placement on
* imbalanced filesystems.
*/
if (unlikely(nfound < (lod_comp->llc_cstripe_count ?
stripe_count : stripe_count_min))) {
/*
* when the decision to use weighted algorithm was made
* we had enough appropriate OSPs, but this state can
* change anytime (no space on OST, broken connection, etc)
* so it's possible OSP won't be able to provide us with
* an object due to just changed state
*/
CDEBUG(D_OTHER, "%s: wanted %d objects, found only %d\n",
lod2obd(lod)->obd_name, stripe_count, nfound);
for (i = 0; i < nfound; i++) {
LASSERT(stripe[i] != NULL);
dt_object_put(env, stripe[i]);
stripe[i] = NULL;
}
/* makes sense to rebalance next time */
set_bit(LQ_DIRTY, &lod->lod_ost_descs.ltd_qos.lq_flags);
clear_bit(LQ_SAME_SPACE, &lod->lod_ost_descs.ltd_qos.lq_flags);
rc = -EAGAIN;
} else if (nfound < lod_comp->llc_stripe_count) {
lod_comp->llc_stripe_count = nfound;
}
/* If there are enough OSTs, a component with overstriping requessted
* will not actually end up overstriped. The comp should reflect this.
*/
if (rc == 0 && !overstriped)
lod_comp->llc_pattern &= ~LOV_PATTERN_OVERSTRIPING;
out:
up_write(&lod->lod_ost_descs.ltd_qos.lq_rw_sem);
out_nolock:
if (pool != NULL) {
up_read(&pool_tgt_rw_sem(pool));
/* put back ref got by lod_find_pool() */
lod_pool_putref(pool);
}
RETURN(rc);
}
/**
* lod_mdt_alloc_qos() - Allocate a striping using an algorithm with weights.
* @env: execution environment for this thread
* @lo: LOD object
* @stripe_idx: starting stripe index to allocate, if it's not 0,
* we are restriping directory
* @stripe_count: total stripe count
* @stripes: striping created
*
* The function allocates remote MDT objects to create a striping, the first
* object was already allocated on current MDT to ensure master object and
* the first object are on the same MDT. The algorithm used is based on weights
* (both free space and inodes), and it's trying to ensure the space/inodes are
* used evenly by MDTs and MDSs. The striping configuration (# of stripes,
* offset, pool) is taken from the object and is prepared by the caller.
*
* If prepared configuration can't be met due to too few MDTs, then allocation
* fails.
*
* No concurrent allocation is allowed on the object and this must be ensured
* by the caller. All the internal structures are protected by the function.
*
* The algorithm has two steps: find available MDTs and calculate their
* weights, then select the MDTs with their weights used as the probability.
* An MDT with a higher weight is proportionately more likely to be selected
* than one with a lower weight.
*
* Return:
* * %positive stripes allocated, and it should be equal to
* lo->ldo_dir_stripe_count
* * %-EAGAIN not enough tgts are found for specified stripe count
* * %-EINVAL requested MDT index is invalid
* * %negative errno on failure
*/
int lod_mdt_alloc_qos(const struct lu_env *env, struct lod_object *lo,
struct dt_object **stripes, u32 stripe_idx,
u32 stripe_count)
{
struct lod_device *lod = lu2lod_dev(lo->ldo_obj.do_lu.lo_dev);
struct lu_tgt_descs *ltd = &lod->lod_mdt_descs;
struct lu_object_conf conf = { .loc_flags = LOC_F_NEW };
struct lu_fid fid = { 0 };
const struct lu_tgt_pool *pool;
struct lu_tgt_desc *mdt;
struct dt_object *dto;
u64 total_weight = 0;
u32 saved_idx = stripe_idx;
u32 mdt_idx;
unsigned int good_mdts;
unsigned int i;
int rc = 0;
ENTRY;
/* Totally skip qos part when qos_threshold_rr=100% */
if (ltd->ltd_qos.lq_threshold_rr == QOS_THRESHOLD_MAX)
return -EAGAIN;
LASSERT(stripe_idx <= stripe_count);
if (stripe_idx == stripe_count)
RETURN(stripe_count);
/* we do not use qos for overstriping, since it will always use all the
* MDTs. So we check if it's truly needed, falling back to rr if it is,
* and otherwise we remove the flag and continue
*/
if (lo->ldo_dir_hash_type & LMV_HASH_FLAG_OVERSTRIPED) {
if (stripe_count > lod->lod_remote_mdt_count + 1)
RETURN(-EAGAIN);
lo->ldo_dir_hash_type &= ~LMV_HASH_FLAG_OVERSTRIPED;
}
/* use MDT pool in @ltd, once MDT pool is supported in the future, it
* can be passed in as argument like OST object allocation.
*/
pool = <d->ltd_tgt_pool;
/* Detect -EAGAIN early, before expensive lock is taken. */
if (!ltd_qos_is_usable(ltd))
RETURN(-EAGAIN);
rc = lod_qos_mdt_in_use_init(env, ltd, stripe_idx, stripe_count, pool,
stripes);
if (rc)
RETURN(rc);
/* Do actual allocation, use write lock here. */
down_write(<d->ltd_qos.lq_rw_sem);
/*
* Check again, while we were sleeping on @lq_rw_sem things could
* change.
*/
if (!ltd_qos_is_usable(ltd))
GOTO(unlock, rc = -EAGAIN);
rc = ltd_qos_penalties_calc(ltd);
if (rc)
GOTO(unlock, rc);
good_mdts = 0;
/* Find all the MDTs that are valid stripe candidates */
for (i = 0; i < pool->op_count; i++) {
if (!test_bit(pool->op_array[i], ltd->ltd_tgt_bitmap))
continue;
mdt = LTD_TGT(ltd, pool->op_array[i]);
mdt->ltd_qos.ltq_usable = 0;
if (mdt->ltd_discon || lod_statfs_check(ltd, mdt))
continue;
if (mdt->ltd_statfs.os_state &
(OS_STATFS_DEGRADED | OS_STATFS_NOCREATE))
continue;
mdt->ltd_qos.ltq_usable = 1;
lu_tgt_qos_weight_calc(mdt, true);
total_weight += mdt->ltd_qos.ltq_weight;
good_mdts++;
}
CDEBUG(D_OTHER, "found %d good MDTs\n", good_mdts);
if (good_mdts < stripe_count - stripe_idx)
GOTO(unlock, rc = -EAGAIN);
/* Find enough MDTs with weighted random allocation. */
while (stripe_idx < stripe_count) {
u64 rand, cur_weight;
cur_weight = 0;
rc = -ENOSPC;
rand = lu_prandom_u64_max(total_weight);
/* On average, this will hit larger-weighted MDTs more often.
* 0-weight MDT will always get used last (only when rand=0) */
for (i = 0; i < pool->op_count; i++) {
int rc2;
mdt_idx = pool->op_array[i];
mdt = LTD_TGT(ltd, mdt_idx);
if (!mdt->ltd_qos.ltq_usable)
continue;
cur_weight += mdt->ltd_qos.ltq_weight;
CDEBUG(D_OTHER, "stripe_count=%d stripe_index=%d cur_weight=%llu rand=%llu total_weight=%llu\n",
stripe_count, stripe_idx, cur_weight, rand,
total_weight);
if (cur_weight < rand)
continue;
CDEBUG(D_OTHER, "stripe=%d to idx=%d\n",
stripe_idx, mdt_idx);
if (lod_qos_is_tgt_used(env, mdt_idx, stripe_idx))
continue;
rc2 = dt_fid_alloc(env, mdt->ltd_tgt, &fid, NULL, NULL);
if (rc2 < 0) {
CDEBUG(D_OTHER, "can't alloc FID on #%u: %d\n",
mdt_idx, rc2);
continue;
}
conf.loc_flags = LOC_F_NEW;
dto = dt_locate_at(env, mdt->ltd_tgt, &fid,
lo->ldo_obj.do_lu.lo_dev->ld_site->ls_top_dev,
&conf);
if (IS_ERR(dto)) {
CDEBUG(D_OTHER, "can't alloc stripe on #%u: %d\n",
mdt_idx, (int) PTR_ERR(dto));
continue;
}
lod_qos_tgt_in_use(env, stripe_idx, mdt_idx);
stripes[stripe_idx] = dto;
ltd_qos_update(ltd, mdt, &total_weight);
stripe_idx++;
rc = 0;
break;
}
/* no MDT found on this iteration, give up */
if (rc)
break;
}
if (unlikely(stripe_idx != stripe_count)) {
/*
* when the decision to use weighted algorithm was made
* we had enough appropriate OSPs, but this state can
* change anytime (no space on MDT, broken connection, etc)
* so it's possible OSP won't be able to provide us with
* an object due to just changed state
*/
CDEBUG(D_OTHER, "%s: wanted %d objects, found only %d\n",
lod2obd(lod)->obd_name, stripe_count, stripe_idx);
for (i = saved_idx; i < stripe_idx; i++) {
LASSERT(stripes[i] != NULL);
dt_object_put(env, stripes[i]);
stripes[i] = NULL;
}
/* makes sense to rebalance next time */
set_bit(LQ_DIRTY, <d->ltd_qos.lq_flags);
clear_bit(LQ_SAME_SPACE, <d->ltd_qos.lq_flags);
rc = -EAGAIN;
} else {
rc = stripe_idx;
}
unlock:
up_write(<d->ltd_qos.lq_rw_sem);
RETURN(rc);
}
/**
* lod_get_stripe_count_plain() - Check stripe count the caller can use.
* @lod: LOD device
* @lo: The lod_object
* @stripe_count: count the caller would like to use
* @overstriping: if overstriping is allowed (overstriping is allowing more
* stripes than available target)
* @flags: Indicates if user specifed stripe count is to be used for default
*
* For new layouts (no initialized components), check the total size of the
* layout against the maximum EA size from the backing file system. This
* stops us from creating a layout which will be too large once initialized.
*
* For existing layouts (with initialized components):
* Find the maximal possible stripe count not greater than \a stripe_count.
* If the provided stripe count is 0, then the filesystem's default is used.
*
* Returns the maximum usable stripe count
*/
__u16 lod_get_stripe_count_plain(struct lod_device *lod, struct lod_object *lo,
__u16 stripe_count, bool overstriping,
enum lod_uses_hint *flags)
{
struct lov_desc *lov_desc = &lod->lod_ost_descs.ltd_lov_desc;
if (!stripe_count)
stripe_count = lov_desc->ld_default_stripe_count;
/* Overstriping allows more stripes than targets */
if (stripe_count > lov_desc->ld_active_tgt_count) {
if (!overstriping) {
*flags |= LOD_USES_DEFAULT_STRIPE;
if (stripe_count == LOV_ALL_STRIPES &&
lod->lod_max_stripecount)
stripe_count = lod->lod_max_stripecount;
else
stripe_count = lov_desc->ld_active_tgt_count;
} else if (stripe_count <= LOV_ALL_STRIPES &&
stripe_count >= LOV_ALL_STRIPES_WIDE) {
stripe_count = lov_desc->ld_active_tgt_count *
(LOV_ALL_STRIPES - stripe_count + 1);
}
}
if (!stripe_count)
stripe_count = 1;
if (stripe_count > LOV_MAX_STRIPE_COUNT)
stripe_count = LOV_MAX_STRIPE_COUNT;
return stripe_count;
}
__u16 lod_get_stripe_count(struct lod_device *lod, struct lod_object *lo,
int comp_idx, __u16 stripe_count, bool overstriping,
enum lod_uses_hint *flags)
{
__u32 max_stripes = LOV_MAX_STRIPE_COUNT_OLD;
/* max stripe count is based on OSD ea size */
unsigned int easize = lod->lod_osd_max_easize;
int i;
ENTRY;
stripe_count = lod_get_stripe_count_plain(lod, lo, stripe_count,
overstriping, flags);
if (lo->ldo_is_composite) {
struct lod_layout_component *lod_comp;
unsigned int header_sz = sizeof(struct lov_comp_md_v1);
unsigned int init_comp_sz = 0;
unsigned int total_comp_sz = 0;
unsigned int comp_sz;
header_sz += sizeof(struct lov_comp_md_entry_v1) *
lo->ldo_comp_cnt;
for (i = 0; i < lo->ldo_comp_cnt; i++) {
unsigned int stripes;
if (i == comp_idx)
continue;
lod_comp = &lo->ldo_comp_entries[i];
/* Extension comp is never inited - 0 stripes on disk */
stripes = lod_comp->llc_flags & LCME_FL_EXTENSION ? 0 :
lod_comp->llc_stripe_count;
comp_sz = lov_mds_md_size(stripes, LOV_MAGIC_V3);
total_comp_sz += comp_sz;
if (lod_comp->llc_flags & LCME_FL_INIT)
init_comp_sz += comp_sz;
}
if (init_comp_sz > 0)
total_comp_sz = init_comp_sz;
header_sz += total_comp_sz;
if (easize > header_sz)
easize -= header_sz;
else
easize = 0;
}
max_stripes = lov_mds_md_max_stripe_count(easize, LOV_MAGIC_V3);
max_stripes = (max_stripes == 0) ? 0 : max_stripes - 1;
stripe_count = min_t(__u16, stripe_count, max_stripes);
RETURN(stripe_count);
}
/**
* lod_use_defined_striping() - Create in-core respresentation for a
* fully-defined striping
* @env: execution environment for this thread
* @mo: LOD object
* @buf: buffer containing the striping
*
* When the caller passes a fully-defined striping (i.e. everything including
* OST object FIDs are defined), then we still need to instantiate LU-cache
* with the objects representing the stripes defined. This function completes
* that task.
*
* Return:
* * %0 on success
* * %negative negated errno on error
*/
int lod_use_defined_striping(const struct lu_env *env,
struct lod_object *mo,
const struct lu_buf *buf)
{
struct lod_layout_component *lod_comp;
struct lov_mds_md_v1 *v1 = buf->lb_buf;
struct lov_mds_md_v3 *v3 = buf->lb_buf;
struct lov_comp_md_v1 *comp_v1 = NULL;
struct lov_ost_data_v1 *objs;
__u32 magic;
__u16 comp_cnt;
__u16 mirror_cnt;
int rc = 0, i;
ENTRY;
mutex_lock(&mo->ldo_layout_mutex);
lod_striping_free_nolock(env, mo);
magic = le32_to_cpu(v1->lmm_magic) & ~LOV_MAGIC_DEFINED;
if (magic != LOV_MAGIC_V1 && magic != LOV_MAGIC_V3 &&
magic != LOV_MAGIC_COMP_V1 && magic != LOV_MAGIC_FOREIGN)
GOTO(unlock, rc = -EINVAL);
/* layout generation must be preserved, otherwise client won't
* be able to refresh it having a more recent generation */
if (magic == LOV_MAGIC_V1 || magic == LOV_MAGIC_V3) {
mo->ldo_is_composite = 0;
comp_cnt = 1;
mirror_cnt = 0;
mo->ldo_layout_gen = le16_to_cpu(v1->lmm_layout_gen);
} else if (magic == LOV_MAGIC_COMP_V1) {
comp_v1 = buf->lb_buf;
comp_cnt = le16_to_cpu(comp_v1->lcm_entry_count);
if (comp_cnt == 0)
GOTO(unlock, rc = -EINVAL);
mirror_cnt = le16_to_cpu(comp_v1->lcm_mirror_count) + 1;
mo->ldo_flr_state = le16_to_cpu(comp_v1->lcm_flags) &
LCM_FL_FLR_MASK;
mo->ldo_is_composite = 1;
mo->ldo_layout_gen = le32_to_cpu(comp_v1->lcm_layout_gen);
} else if (magic == LOV_MAGIC_FOREIGN) {
struct lov_foreign_md *foreign;
size_t length;
if (buf->lb_len < offsetof(typeof(*foreign), lfm_value)) {
CDEBUG(D_LAYOUT,
"buf len %zu < min lov_foreign_md size (%zu)\n",
buf->lb_len,
offsetof(typeof(*foreign), lfm_value));
GOTO(out, rc = -EINVAL);
}
foreign = (struct lov_foreign_md *)buf->lb_buf;
length = lov_foreign_size_le(foreign);
if (buf->lb_len < length) {
CDEBUG(D_LAYOUT,
"buf len %zu < this lov_foreign_md size (%zu)\n",
buf->lb_len, length);
GOTO(out, rc = -EINVAL);
}
/* just cache foreign LOV EA raw */
rc = lod_alloc_foreign_lov(mo, length);
if (rc)
GOTO(out, rc);
memcpy(mo->ldo_foreign_lov, buf->lb_buf, length);
GOTO(out, rc);
} else {
GOTO(out, rc = -EINVAL);
}
rc = lod_alloc_comp_entries(mo, mirror_cnt, comp_cnt);
if (rc)
GOTO(unlock, rc);
for (i = 0; i < comp_cnt; i++) {
struct lu_extent *ext;
char *pool_name;
__u32 offs;
lod_comp = &mo->ldo_comp_entries[i];
if (mo->ldo_is_composite) {
struct lov_comp_md_entry_v1 *ent =
&comp_v1->lcm_entries[i];
offs = le32_to_cpu(ent->lcme_offset);
v1 = (struct lov_mds_md_v1 *)((char *)comp_v1 + offs);
v3 = (struct lov_mds_md_v3 *)v1;
magic = le32_to_cpu(v1->lmm_magic);
ext = &ent->lcme_extent;
lod_comp->llc_extent.e_start =
le64_to_cpu(ext->e_start);
lod_comp->llc_extent.e_end = le64_to_cpu(ext->e_end);
lod_comp->llc_flags = le32_to_cpu(ent->lcme_flags);
lod_comp->llc_timestamp = lcme_timestamp_time_unpack(
le64_to_cpu(ent->lcme_time_and_id));
lod_comp->llc_mirror_link_id = lcme_timestamp_id_unpack(
le64_to_cpu(ent->lcme_time_and_id));
lod_comp->llc_dstripe_count =
comp_v1->lcm_entries[i].lcme_dstripe_count;
lod_comp->llc_cstripe_count =
comp_v1->lcm_entries[i].lcme_cstripe_count;
lod_comp->llc_id = le32_to_cpu(ent->lcme_id);
if (lod_comp->llc_id == LCME_ID_INVAL)
GOTO(out, rc = -EINVAL);
lod_comp->llc_magic = magic;
if (magic == LOV_MAGIC_FOREIGN) {
rc = lod_init_comp_foreign(lod_comp, v1);
if (rc)
GOTO(out, rc);
continue;
}
} else {
lod_comp->llc_magic = magic;
}
pool_name = NULL;
if (magic == LOV_MAGIC_V1) {
objs = &v1->lmm_objects[0];
} else if (magic == LOV_MAGIC_V3) {
objs = &v3->lmm_objects[0];
if (v3->lmm_pool_name[0] != '\0')
pool_name = v3->lmm_pool_name;
} else {
CDEBUG(D_LAYOUT, "Invalid magic %x\n", magic);
GOTO(out, rc = -EINVAL);
}
lod_comp->llc_pattern = le32_to_cpu(v1->lmm_pattern);
lod_comp->llc_stripe_size = le32_to_cpu(v1->lmm_stripe_size);
lod_comp->llc_stripe_count = le16_to_cpu(v1->lmm_stripe_count);
/* set parity component pattern */
if (lod_comp->llc_flags & LCME_FL_PARITY)
lod_comp->llc_pattern |= LOV_PATTERN_PARITY;
/**
* limit stripe count so that it's less than/equal to
* extent_size / stripe_size.
*
* Note: extension size reused llc_stripe_size field and
* uninstantiated component could be defined with
* extent_start == extent_end as extension component will
* expand it later.
*/
if (mo->ldo_is_composite &&
!(lod_comp->llc_flags & LCME_FL_EXTENSION) &&
(lod_comp_inited(lod_comp) ||
lod_comp->llc_extent.e_start <
lod_comp->llc_extent.e_end) &&
!(lod_comp->llc_stripe_count <= LOV_ALL_STRIPES &&
lod_comp->llc_stripe_count >= LOV_ALL_STRIPES_WIDE) &&
lod_comp->llc_extent.e_end != OBD_OBJECT_EOF &&
(__u64)lod_comp->llc_stripe_count *
lod_comp->llc_stripe_size >
(lod_comp->llc_extent.e_end - lod_comp->llc_extent.e_start))
lod_comp->llc_stripe_count =
DIV_ROUND_UP(lod_comp->llc_extent.e_end -
lod_comp->llc_extent.e_start,
lod_comp->llc_stripe_size);
lod_comp->llc_layout_gen = le16_to_cpu(v1->lmm_layout_gen);
/**
* The stripe_offset of an uninit-ed component is stored in
* the lmm_layout_gen
*/
if (mo->ldo_is_composite && !lod_comp_inited(lod_comp))
lod_comp->llc_stripe_offset = lod_comp->llc_layout_gen;
lod_obj_set_pool(mo, i, pool_name);
if ((!mo->ldo_is_composite || lod_comp_inited(lod_comp)) &&
!(lod_comp->llc_pattern & LOV_PATTERN_F_RELEASED) &&
!(lod_comp->llc_pattern & LOV_PATTERN_MDT)) {
rc = lod_initialize_objects(env, mo, objs, i);
if (rc)
GOTO(out, rc);
}
}
rc = lod_fill_mirrors(mo);
GOTO(out, rc);
out:
if (rc)
lod_striping_free_nolock(env, mo);
unlock:
mutex_unlock(&mo->ldo_layout_mutex);
RETURN(rc);
}
void lod_qos_set_pool(struct lod_object *lo, int pos, const char *pool_name)
{
struct lod_device *d = lu2lod_dev(lod2lu_obj(lo)->lo_dev);
struct lod_layout_component *lod_comp;
struct lod_pool_desc *pool = NULL;
__u32 idx;
int j, rc = 0;
/* In the function below, .hs_keycmp resolves to
* pool_hashkey_keycmp() */
if (pool_name)
pool = lod_find_pool(d, pool_name);
if (!pool) {
lod_obj_set_pool(lo, pos, pool_name);
return;
}
lod_comp = &lo->ldo_comp_entries[pos];
if (lod_comp->llc_stripe_offset != LOV_OFFSET_DEFAULT) {
if (lod_comp->llc_ostlist.op_count) {
for (j = 0; j < lod_comp->llc_ostlist.op_count; j++) {
idx = lod_comp->llc_ostlist.op_array[j];
rc = lod_check_index_in_pool(idx, pool);
if (rc)
break;
}
} else {
idx = lod_comp->llc_stripe_offset;
rc = lod_check_index_in_pool(idx, pool);
}
if (rc) {
CDEBUG(D_LAYOUT, "%s: index %u is not in the pool %s, "
"dropping the pool\n", lod2obd(d)->obd_name,
idx, pool_name);
pool_name = NULL;
}
}
if (pool_name &&
lod_comp->llc_stripe_count > pool_tgt_count(pool) &&
!(lod_comp->llc_pattern & LOV_PATTERN_OVERSTRIPING))
lod_comp->llc_stripe_count = pool_tgt_count(pool);
lod_pool_putref(pool);
lod_obj_set_pool(lo, pos, pool_name);
}
/**
* lod_qos_parse_config() - Parse suggested striping configuration.
* @env: execution environment for this thread
* @lo: LOD object
* @buf: buffer containing the striping
*
* The caller gets a suggested striping configuration from a number of sources
* including per-directory default and applications. Then it needs to verify
* the suggested striping is valid, apply missing bits and store the resulting
* configuration in the object to be used by the allocator later. Must not be
* called concurrently against the same object. It's OK to provide a
* fully-defined striping.
*
* Return:
* * %0 on success
* * %negative negated errno on error
*/
int lod_qos_parse_config(const struct lu_env *env, struct lod_object *lo,
const struct lu_buf *buf)
{
struct lod_layout_component *lod_comp;
struct lod_device *d = lu2lod_dev(lod2lu_obj(lo)->lo_dev);
struct lov_desc *desc = &d->lod_ost_descs.ltd_lov_desc;
struct lov_user_md_v1 *v1 = NULL;
struct lov_user_md_v3 *v3 = NULL;
struct lov_comp_md_v1 *comp_v1 = NULL;
struct lov_foreign_md *lfm = NULL;
char def_pool[LOV_MAXPOOLNAME + 1];
__u32 magic;
__u16 comp_cnt;
__u16 mirror_cnt;
int i, rc;
ENTRY;
if (buf == NULL || buf->lb_buf == NULL || buf->lb_len == 0)
RETURN(0);
memset(def_pool, 0, sizeof(def_pool));
if (lo->ldo_comp_entries != NULL)
lod_layout_get_pool(lo->ldo_comp_entries, lo->ldo_comp_cnt,
def_pool, sizeof(def_pool));
/* free default striping info */
if (lo->ldo_is_foreign)
lod_free_foreign_lov(lo);
else
lod_free_comp_entries(lo);
rc = lod_verify_striping(env, d, lo, buf, false);
if (rc)
RETURN(-EINVAL);
v3 = buf->lb_buf;
v1 = buf->lb_buf;
comp_v1 = buf->lb_buf;
/* {lmm,lfm}_magic position/length work for all LOV formats */
magic = v1->lmm_magic;
if (unlikely(le32_to_cpu(magic) & LOV_MAGIC_DEFINED)) {
/* try to use as fully defined striping */
rc = lod_use_defined_striping(env, lo, buf);
RETURN(rc);
}
switch (magic) {
case __swab32(LOV_USER_MAGIC_V1):
lustre_swab_lov_user_md_v1(v1);
magic = v1->lmm_magic;
fallthrough;
case LOV_USER_MAGIC_V1:
break;
case __swab32(LOV_USER_MAGIC_V3):
lustre_swab_lov_user_md_v3(v3);
magic = v3->lmm_magic;
fallthrough;
case LOV_USER_MAGIC_V3:
break;
case __swab32(LOV_USER_MAGIC_SPECIFIC):
lustre_swab_lov_user_md_v3(v3);
lustre_swab_lov_user_md_objects(v3->lmm_objects,
v3->lmm_stripe_count);
magic = v3->lmm_magic;
fallthrough;
case LOV_USER_MAGIC_SPECIFIC:
break;
case __swab32(LOV_USER_MAGIC_COMP_V1):
lustre_swab_lov_comp_md_v1(comp_v1);
magic = comp_v1->lcm_magic;
fallthrough;
case LOV_USER_MAGIC_COMP_V1:
break;
case __swab32(LOV_USER_MAGIC_FOREIGN):
lfm = buf->lb_buf;
__swab32s(&lfm->lfm_magic);
__swab32s(&lfm->lfm_length);
__swab32s(&lfm->lfm_type);
__swab32s(&lfm->lfm_flags);
magic = lfm->lfm_magic;
fallthrough;
case LOV_USER_MAGIC_FOREIGN:
if (!lfm)
lfm = buf->lb_buf;
rc = lod_alloc_foreign_lov(lo, lov_foreign_size(lfm));
if (rc)
RETURN(rc);
memcpy(lo->ldo_foreign_lov, buf->lb_buf,
lov_foreign_size(lfm));
RETURN(0);
default:
CERROR("%s: unrecognized magic %X\n",
lod2obd(d)->obd_name, magic);
RETURN(-EINVAL);
}
lustre_print_user_md(D_OTHER, v1, "parse config");
if (magic == LOV_USER_MAGIC_COMP_V1) {
comp_cnt = comp_v1->lcm_entry_count;
if (comp_cnt == 0)
RETURN(-EINVAL);
mirror_cnt = comp_v1->lcm_mirror_count + 1;
if (mirror_cnt > 1)
lo->ldo_flr_state = LCM_FL_RDONLY;
lo->ldo_is_composite = 1;
} else {
comp_cnt = 1;
mirror_cnt = 0;
lo->ldo_is_composite = 0;
}
rc = lod_alloc_comp_entries(lo, mirror_cnt, comp_cnt);
if (rc)
RETURN(rc);
LASSERT(lo->ldo_comp_entries);
for (i = 0; i < comp_cnt; i++) {
struct lu_extent *ext;
char *pool_name;
struct lov_comp_md_entry_v1 *lcme = NULL;
lod_comp = &lo->ldo_comp_entries[i];
if (lo->ldo_is_composite) {
lcme = &comp_v1->lcm_entries[i];
v1 = (struct lov_user_md *)((char *)comp_v1 +
lcme->lcme_offset);
ext = &lcme->lcme_extent;
lod_comp->llc_extent = *ext;
lod_comp->llc_flags = lcme->lcme_flags &
LCME_CL_COMP_FLAGS;
lod_comp->llc_mirror_link_id =
lcme_timestamp_id_unpack(lcme->lcme_time_and_id);
lod_comp->llc_dstripe_count = lcme->lcme_dstripe_count;
lod_comp->llc_cstripe_count = lcme->lcme_cstripe_count;
}
pool_name = NULL;
if (def_pool[0] != '\0')
pool_name = def_pool;
if (v1->lmm_magic == LOV_USER_MAGIC_V3 ||
v1->lmm_magic == LOV_USER_MAGIC_SPECIFIC) {
v3 = (struct lov_user_md_v3 *)v1;
if (lov_pool_is_ignored(v3->lmm_pool_name))
pool_name = NULL;
else if (v3->lmm_pool_name[0] != '\0' &&
!lov_pool_is_inherited(v3->lmm_pool_name))
pool_name = v3->lmm_pool_name;
if (v3->lmm_magic == LOV_USER_MAGIC_SPECIFIC) {
rc = lod_comp_copy_ost_lists(lod_comp, v3);
if (rc)
GOTO(free_comp, rc);
}
}
if (v1->lmm_pattern == 0)
v1->lmm_pattern = LOV_PATTERN_RAID0;
if (!lov_pattern_available(v1->lmm_pattern)) {
CDEBUG(D_LAYOUT, "%s: invalid pattern: %x\n",
lod2obd(d)->obd_name, v1->lmm_pattern);
GOTO(free_comp, rc = -EINVAL);
}
lod_comp->llc_pattern = v1->lmm_pattern;
lod_comp->llc_stripe_size = v1->lmm_stripe_size;
lod_adjust_stripe_size(lod_comp, desc->ld_default_stripe_size);
lod_comp->llc_stripe_count = desc->ld_default_stripe_count;
if (v1->lmm_stripe_count ||
(lov_pattern(v1->lmm_pattern) & LOV_PATTERN_MDT))
lod_comp->llc_stripe_count = v1->lmm_stripe_count;
if ((lov_pattern(lod_comp->llc_pattern) & LOV_PATTERN_MDT) &&
lod_comp->llc_stripe_count != 0) {
CDEBUG(D_LAYOUT, "%s: invalid stripe count: %u\n",
lod2obd(d)->obd_name,
lod_comp->llc_stripe_count);
GOTO(free_comp, rc = -EINVAL);
}
/* set parity component pattern */
if (lod_comp->llc_flags & LCME_FL_PARITY)
lod_comp->llc_pattern |= LOV_PATTERN_PARITY;
/**
* limit stripe count so that it's less than/equal to
* extent_size / stripe_size.
*
* Note: extension size reused llc_stripe_size field and
* uninstantiated component could be defined with
* extent_start == extent_end as extension component will
* expand it later.
*/
if (lo->ldo_is_composite &&
!(lod_comp->llc_flags & LCME_FL_EXTENSION) &&
!(lod_comp->llc_stripe_count <= LOV_ALL_STRIPES &&
lod_comp->llc_stripe_count >= LOV_ALL_STRIPES_WIDE) &&
(lod_comp_inited(lod_comp) ||
lod_comp->llc_extent.e_start <
lod_comp->llc_extent.e_end) &&
lod_comp->llc_extent.e_end != OBD_OBJECT_EOF &&
lod_comp->llc_stripe_count * lod_comp->llc_stripe_size >
(lod_comp->llc_extent.e_end - lod_comp->llc_extent.e_start))
lod_comp->llc_stripe_count =
DIV_ROUND_UP(lod_comp->llc_extent.e_end -
lod_comp->llc_extent.e_start,
lod_comp->llc_stripe_size);
if ((lov_pattern(lod_comp->llc_pattern) & LOV_PATTERN_COMPRESS) &&
lcme) {
lod_comp->llc_compr_type = lcme->lcme_compr_type;
lod_comp->llc_compr_lvl = lcme->lcme_compr_lvl;
lod_comp->llc_compr_chunk_lum_bits =
lcme->lcme_compr_chunk_lum_bits;
}
lod_comp->llc_stripe_offset = v1->lmm_stripe_offset;
lod_qos_set_pool(lo, i, pool_name);
}
RETURN(0);
free_comp:
lod_free_comp_entries(lo);
RETURN(rc);
}
/*
* prepare enough OST avoidance bitmap space
*/
static int lod_prepare_avoidance(const struct lu_env *env,
struct lod_object *lo)
{
struct lod_device *d = lu2lod_dev(lo->ldo_obj.do_lu.lo_dev);
struct lod_avoid_guide *lag = &lod_env_info(env)->lti_avoid;
unsigned long *bitmap = NULL;
__u32 new_ost_size = d->lod_ost_size;
__u32 *new_oss = NULL;
__u32 new_oss_size;
/*
* usually there are multiple OSTs in one OSS, but we don't
* know the exact OSS number, so we choose a safe option,
* using OST count to allocate the array to store the OSS
* id.
*/
new_oss_size = d->lod_ost_count;
/* reset OSS avoid guide array */
lag->lag_oaa_count = 0;
if (lag->lag_oss_avoid_array && lag->lag_oaa_size < d->lod_ost_count) {
OBD_FREE_PTR_ARRAY(lag->lag_oss_avoid_array, lag->lag_oaa_size);
lag->lag_oss_avoid_array = NULL;
lag->lag_oaa_size = 0;
}
/* init OST avoid guide bitmap */
lag->lag_ost_avail = d->lod_ost_count;
if (lag->lag_ost_avoid_bitmap) {
if (new_ost_size <= lag->lag_ost_avoid_size) {
bitmap_zero(lag->lag_ost_avoid_bitmap,
lag->lag_ost_avoid_size);
} else {
bitmap_free(lag->lag_ost_avoid_bitmap);
lag->lag_ost_avoid_bitmap = NULL;
}
}
if (!lag->lag_ost_avoid_bitmap) {
bitmap = bitmap_zalloc(new_ost_size, GFP_KERNEL);
if (!bitmap)
return -ENOMEM;
}
if (!lag->lag_oss_avoid_array) {
OBD_ALLOC_PTR_ARRAY(new_oss, new_oss_size);
if (!new_oss) {
bitmap_free(bitmap);
return -ENOMEM;
}
}
if (new_oss) {
lag->lag_oss_avoid_array = new_oss;
lag->lag_oaa_size = new_oss_size;
}
if (bitmap) {
lag->lag_ost_avoid_bitmap = bitmap;
lag->lag_ost_avoid_size = new_ost_size;
}
return 0;
}
/*
* Collect information of used OSTs and OSSs in the overlapped components
* of other mirrors
*/
static void lod_collect_avoidance(struct lod_object *lo,
struct lod_avoid_guide *lag,
int comp_idx)
{
struct lod_device *lod = lu2lod_dev(lo->ldo_obj.do_lu.lo_dev);
struct lod_layout_component *lod_comp = &lo->ldo_comp_entries[comp_idx];
unsigned long *bitmap = lag->lag_ost_avoid_bitmap;
int i, j;
/* iterate components */
for (i = 0; i < lo->ldo_comp_cnt; i++) {
struct lod_layout_component *comp;
/**
* skip mirror containing component[comp_idx], we only
* collect OSTs info of conflicting component in other mirrors,
* so that during read, if OSTs of a mirror's component are
* not available, we still have other mirror with different
* OSTs to read the data.
*/
comp = &lo->ldo_comp_entries[i];
if (comp->llc_id != LCME_ID_INVAL &&
mirror_id_of(comp->llc_id) ==
mirror_id_of(lod_comp->llc_id))
continue;
/**
* skip non-overlapped or un-instantiated components,
* NOTE: don't use lod_comp_inited(comp) to judge
* whether @comp has been inited, since during
* declare phase, comp->llc_stripe has been allocated
* while it's init flag not been set until the exec
* phase.
*/
if (!lu_extent_is_overlapped(&comp->llc_extent,
&lod_comp->llc_extent) ||
!comp->llc_stripe)
continue;
/**
* collect used OSTs index and OSS info from a
* component
*/
for (j = 0; j < comp->llc_stripe_count; j++) {
struct lod_tgt_desc *ost;
struct lu_svr_qos *lsq;
int k;
ost = OST_TGT(lod, comp->llc_ost_indices[j]);
lsq = ost->ltd_qos.ltq_svr;
if (test_bit(ost->ltd_index, bitmap))
continue;
CDEBUG(D_OTHER, "OST%d used in conflicting mirror component\n", ost->ltd_index);
set_bit(ost->ltd_index, bitmap);
lag->lag_ost_avail--;
for (k = 0; k < lag->lag_oaa_count; k++) {
if (lag->lag_oss_avoid_array[k] ==
lsq->lsq_id)
break;
}
if (k == lag->lag_oaa_count) {
lag->lag_oss_avoid_array[k] =
lsq->lsq_id;
lag->lag_oaa_count++;
}
}
}
}
/**
* lod_comp_find_linked() - Find the data/parity component linked to @comp.
* @lo: LOD object
* @comp: an EC data or parity component
*
* A data component and its parity component are paired via their
* "mirror_link_id". It is either a link id (both comps share the same id) or
* a mirror id (each comp's "mirror_link_id" points to the other's mirror id);
* the llapi initially assigns the link id and lod_bind_data_parity() later
* resolves it to the permanent mirror id. Linked comps cover the same extent.
*
* Return: the linked component of the opposite kind, or NULL if none found.
*/
static struct lod_layout_component *
lod_comp_find_linked(struct lod_object *lo, struct lod_layout_component *comp)
{
bool is_link_id = comp->llc_flags & LCME_FL_IS_LINK_ID;
bool is_parity = comp->llc_flags & LCME_FL_PARITY;
int j;
for (j = 0; j < lo->ldo_comp_cnt; j++) {
struct lod_layout_component *tmp = &lo->ldo_comp_entries[j];
bool tmp_is_parity = tmp->llc_flags & LCME_FL_PARITY;
__u16 link_id;
/* the linked comp is of the opposite kind (data/parity) */
if (is_parity == tmp_is_parity)
continue;
/* resolve the candidate's link id for the comparison */
if (is_link_id) {
if (!(tmp->llc_flags & LCME_FL_IS_LINK_ID))
continue;
link_id = tmp->llc_mirror_link_id;
} else {
link_id = mirror_id_of(tmp->llc_id);
}
/* linked comps share a link id and cover the same extent */
if (link_id == comp->llc_mirror_link_id &&
lu_extent_is_equal(&tmp->llc_extent, &comp->llc_extent))
return tmp;
}
return NULL;
}
/**
* lod_comp_prep_create() - Create a striping for an object.
* @env: execution environment for this thread
* @lo: LOD object
* @attr: attributes OST objects will be declared with
* @th: transaction handle
* @comp_idx: index of ldo_comp_entries
* @reserve: space to reserve on target device
*
* The function creates a new striping for the object. The function tries QoS
* algorithm first unless free space is distributed evenly among OSTs, but
* by default RR algorithm is preferred due to internal concurrency (QoS is
* serialized). The caller must ensure no concurrent calls to the function
* are made against the same object.
*
* Return:
* * %0 on success
* * %negative negated errno on error
*/
int lod_comp_prep_create(const struct lu_env *env, struct lod_object *lo,
struct lu_attr *attr, struct thandle *th,
int comp_idx, __u64 reserve)
{
struct lod_layout_component *lod_comp, *data_comp = NULL;
struct lod_device *d = lu2lod_dev(lod2lu_obj(lo)->lo_dev);
struct lod_avoid_guide *lag = &lod_env_info(env)->lti_avoid;
struct dt_object **stripe = NULL;
__u32 *ost_indices = NULL;
enum lod_uses_hint flags = LOD_USES_ASSIGNED_STRIPE;
int stripe_len;
int i, rc = 0;
bool is_parity;
struct ec_split_comp sc = {0};
ENTRY;
LASSERT(lo);
if (IS_ERR_OR_NULL(d))
RETURN(d ? PTR_ERR(d) : -EINVAL);
LASSERT(lo->ldo_comp_cnt > comp_idx && lo->ldo_comp_entries != NULL);
lod_comp = &lo->ldo_comp_entries[comp_idx];
LASSERT(!(lod_comp->llc_flags & LCME_FL_EXTENSION));
/* A foreign/HSM component is being created */
if (lod_comp->llc_magic == LOV_MAGIC_FOREIGN)
RETURN(0);
/* A released component is being created */
if (lod_comp->llc_pattern & LOV_PATTERN_F_RELEASED)
RETURN(0);
/* A Data-on-MDT component is being created */
if (lov_pattern(lod_comp->llc_pattern) & LOV_PATTERN_MDT)
RETURN(0);
if (lod_comp->llc_pool)
lod_check_and_spill_pool(env, d, &lod_comp->llc_pool);
/*
* A parity or EC-protected data component is paired with its linked
* component (see lod_comp_find_linked()). For a parity we size it (one
* OST per parity per raid set) and remember the data component for the
* raidset-aware allocator; for both we compute the raid-set split (sc)
* the allocators need.
*/
is_parity = lod_comp->llc_flags & LCME_FL_PARITY;
if (is_parity || lod_comp->llc_dstripe_count) {
struct lod_layout_component *linked, *data, *parity;
/* a parity holds cstripe_count stripes per raid set */
if (is_parity)
lod_comp->llc_stripe_count =
lod_comp->llc_cstripe_count;
if (lod_comp->llc_mirror_link_id == 0)
GOTO(out, rc = -EINVAL);
linked = lod_comp_find_linked(lo, lod_comp);
/* a parity component must have a matching data component */
if (is_parity && !linked)
GOTO(out, rc = -EINVAL);
if (linked) {
data = is_parity ? linked : lod_comp;
parity = is_parity ? lod_comp : linked;
/*
* Raid-set split from the data stripe count and the
* parity's data width; the allocators use it to lay
* out both components.
*/
ec_split_stripes(data->llc_stripe_count,
parity->llc_dstripe_count, &sc);
if (is_parity) {
/* one OST per raid set and parity */
data_comp = data;
lod_comp->llc_stripe_count =
(sc.esc_n0 + sc.esc_n1) *
lod_comp->llc_cstripe_count;
} else if (data->llc_pattern &
LOV_PATTERN_OVERSTRIPING) {
/* TODO: should be set from userspace */
parity->llc_pattern |= LOV_PATTERN_OVERSTRIPING;
}
}
}
if (likely(!lod_comp->llc_stripe)) {
/*
* no striping has been created so far
*/
LASSERT(lod_comp->llc_stripe_count);
/*
* statfs and check OST targets now, since ld_active_tgt_count
* could be changed if some OSTs are [de]activated manually.
*/
lod_qos_statfs_update(env, d, &d->lod_ost_descs);
/*
* For EC components keep the requested stripe count (the
* RR/QoS paths may shrink geometry if OSTs are short).
* Only reject here when the LOV EA cannot hold that many
* stripes. lod_get_stripe_count() normally also clamps to
* ld_active_tgt_count when overstriping is false; pass
* overstriping=true so max_stripes is the EA capacity, not
* the OST count. A plain -c 24 on 16 OSTs is clamped; an
* EC -c 24 must not hard-fail with -E2BIG for the same.
* Use a local flags so LOD_USES_DEFAULT_STRIPE is not set
* on the caller's flags as a side effect.
*/
if (lod_comp->llc_cstripe_count) {
__u16 max_stripes;
enum lod_uses_hint ea_flags =
LOD_USES_ASSIGNED_STRIPE;
max_stripes = lod_get_stripe_count(d, lo,
comp_idx, LOV_MAX_STRIPE_COUNT,
true, &ea_flags);
if (lod_comp->llc_stripe_count > max_stripes) {
CDEBUG(D_OTHER,
"EC stripe count %u exceeds LOV EA capacity %u\n",
lod_comp->llc_stripe_count,
max_stripes);
GOTO(out, rc = -E2BIG);
}
stripe_len = lod_comp->llc_stripe_count;
} else {
stripe_len = lod_get_stripe_count(d, lo, comp_idx,
lod_comp->llc_stripe_count,
lod_comp->llc_pattern &
LOV_PATTERN_OVERSTRIPING,
&flags);
}
if (stripe_len == 0)
GOTO(out, rc = -ERANGE);
lod_comp->llc_stripe_count = stripe_len;
OBD_ALLOC_PTR_ARRAY(stripe, stripe_len);
if (stripe == NULL)
GOTO(out, rc = -ENOMEM);
OBD_ALLOC_PTR_ARRAY(ost_indices, stripe_len);
if (!ost_indices)
GOTO(out, rc = -ENOMEM);
repeat:
lod_getref(&d->lod_ost_descs);
/* XXX: support for non-0 files w/o objects */
CDEBUG(D_OTHER, "tgt_count %d stripe_count %d\n",
d->lod_ost_count, stripe_len);
if (lod_comp->llc_ostlist.op_array &&
lod_comp->llc_ostlist.op_count) {
rc = lod_alloc_ost_list(env, lo, stripe, ost_indices,
th, comp_idx, reserve);
} else if (lod_comp->llc_stripe_offset == LOV_OFFSET_DEFAULT) {
/**
* collect OSTs and OSSs used in other mirrors whose
* components cross the ldo_comp_entries[comp_idx]
*/
rc = lod_prepare_avoidance(env, lo);
if (rc)
GOTO(put_ldts, rc);
CDEBUG(D_OTHER, "collecting conflict osts for comp[%d]\n",
comp_idx);
lod_collect_avoidance(lo, lag, comp_idx);
rc = lod_ost_alloc_qos(env, lo, stripe, ost_indices,
flags, th, comp_idx, reserve,
&sc, data_comp);
if (rc == -EAGAIN)
rc = lod_ost_alloc_rr(env, lo, stripe,
ost_indices, flags, th,
comp_idx, reserve, &sc,
data_comp);
} else {
rc = lod_ost_alloc_specific(env, lo, stripe,
ost_indices, flags, th,
comp_idx, reserve);
/*
* For uninitialized components with invalid stripe
* offset, fall back to QoS allocation
*/
if (rc == -EAGAIN) {
rc = lod_ost_alloc_qos(env, lo, stripe,
ost_indices, flags, th,
comp_idx, reserve,
&sc, data_comp);
if (rc == -EAGAIN)
rc = lod_ost_alloc_rr(env, lo, stripe,
ost_indices,
flags, th,
comp_idx,
reserve, &sc,
data_comp);
}
}
put_ldts:
lod_putref(d, &d->lod_ost_descs);
if (rc < 0) {
for (i = 0; i < stripe_len; i++)
if (stripe[i] != NULL)
dt_object_put(env, stripe[i]);
/* In case there is no space on any OST, let's ignore
* the @reserve space to avoid an error at the init
* time, probably the actual IO will be less than the
* given @reserve space (aka extension_size). */
if (reserve) {
reserve = 0;
goto repeat;
}
lod_comp->llc_stripe_count = 0;
} else {
lod_comp->llc_layout_gen = 0;
lod_comp->llc_stripe = stripe;
lod_comp->llc_ost_indices = ost_indices;
lod_comp->llc_stripes_allocated = stripe_len;
}
} else {
/*
* lod_qos_parse_config() found supplied buf as a predefined
* striping (not a hint), so it allocated all the object
* now we need to create them
*/
for (i = 0; i < lod_comp->llc_stripe_count; i++) {
struct dt_object *o;
o = lod_comp->llc_stripe[i];
LASSERT(o);
rc = lod_sub_declare_create(env, o, attr, NULL,
NULL, th);
if (rc < 0) {
CERROR("can't declare create: %d\n", rc);
break;
}
}
/**
* Clear LCME_FL_INIT for the component so that
* lod_striping_create() can create the striping objects
* in replay.
*/
lod_comp_unset_init(lod_comp);
}
out:
if (rc < 0) {
if (stripe)
OBD_FREE_PTR_ARRAY(stripe, stripe_len);
if (ost_indices)
OBD_FREE_PTR_ARRAY(ost_indices, stripe_len);
}
RETURN(rc);
}
int lod_prepare_create(const struct lu_env *env, struct lod_object *lo,
struct lu_attr *attr, const struct lu_buf *buf,
struct thandle *th)
{
struct lod_device *d = lu2lod_dev(lod2lu_obj(lo)->lo_dev);
uint64_t size = 0;
int i;
int rc;
ENTRY;
LASSERT(lo);
/* no OST available */
/* XXX: should we be waiting a bit to prevent failures during
* cluster initialization? */
if (!d->lod_ost_count)
RETURN(-EIO);
/*
* by this time, the object's ldo_stripe_count and ldo_stripe_size
* contain default value for striping: taken from the parent
* or from filesystem defaults
*
* in case the caller is passing lovea with new striping config,
* we may need to parse lovea and apply new configuration
*/
rc = lod_qos_parse_config(env, lo, buf);
if (rc)
RETURN(rc);
if (attr->la_valid & LA_SIZE)
size = attr->la_size;
/**
* prepare OST object creation for the component covering file's
* size, the 1st component (including plain layout file) is always
* instantiated.
*/
for (i = 0; i < lo->ldo_comp_cnt; i++) {
struct lod_layout_component *lod_comp;
struct lu_extent *extent;
lod_comp = &lo->ldo_comp_entries[i];
extent = &lod_comp->llc_extent;
CDEBUG(D_OTHER, "comp[%d] %lld "DEXT"\n", i, size, PEXT(extent));
if (!lo->ldo_is_composite || size >= extent->e_start) {
rc = lod_comp_prep_create(env, lo, attr, th, i, 0);
if (rc)
break;
}
}
RETURN(rc);
}
