fs/zfs/vdev_mirror.c

	vdev_mirror.c revision fa9e4066f08beec538e775443c5be79dd423fcab
/*
 * CDDL HEADER START
 *
 * The contents of this file are subject to the terms of the
 * Common Development and Distribution License, Version 1.0 only
 * (the "License").  You may not use this file except in compliance
 * with the License.
 *
 * You can obtain a copy of the license at usr/src/OPENSOLARIS.LICENSE
 * or http://www.opensolaris.org/os/licensing.
 * See the License for the specific language governing permissions
 * and limitations under the License.
 *
 * When distributing Covered Code, include this CDDL HEADER in each
 * file and include the License file at usr/src/OPENSOLARIS.LICENSE.
 * If applicable, add the following below this CDDL HEADER, with the
 * fields enclosed by brackets "[]" replaced with your own identifying
 * information: Portions Copyright [yyyy] [name of copyright owner]
 *
 * CDDL HEADER END
 */
/*
 * Copyright 2005 Sun Microsystems, Inc.  All rights reserved.
 * Use is subject to license terms.
 */

#pragma ident   "%Z%%M% %I% %E% SMI"

#include <sys/zfs_context.h>
#include <sys/spa.h>
#include <sys/vdev_impl.h>
#include <sys/zio.h>
#include <sys/fs/zfs.h>

/*
 * Virtual device vector for mirroring.
 */

typedef struct mirror_map {
    int mm_error;
    short   mm_tried;
    short   mm_skipped;
} mirror_map_t;

static mirror_map_t *
vdev_mirror_map_alloc(zio_t *zio)
{
    zio->io_vsd = kmem_zalloc(zio->io_vd->vdev_children *
        sizeof (mirror_map_t), KM_SLEEP);
    return (zio->io_vsd);
}

static void
vdev_mirror_map_free(zio_t *zio)
{
    kmem_free(zio->io_vsd,
        zio->io_vd->vdev_children * sizeof (mirror_map_t));
    zio->io_vsd = NULL;
}

static int
vdev_mirror_open(vdev_t *vd, uint64_t *asize, uint64_t *ashift)
{
    vdev_t *cvd;
    uint64_t c;
    int numerrors = 0;
    int ret, lasterror = 0;

    if (vd->vdev_children == 0) {
        vd->vdev_stat.vs_aux = VDEV_AUX_BAD_LABEL;
        return (EINVAL);
    }

    for (c = 0; c < vd->vdev_children; c++) {
        cvd = vd->vdev_child[c];

        if ((ret = vdev_open(cvd)) != 0) {
            lasterror = ret;
            numerrors++;
            continue;
        }

        *asize = MIN(*asize - 1, cvd->vdev_asize - 1) + 1;
        *ashift = cvd->vdev_ashift;
    }

    if (numerrors == vd->vdev_children) {
        vd->vdev_stat.vs_aux = VDEV_AUX_NO_REPLICAS;
        return (lasterror);
    }

    return (0);
}

static void
vdev_mirror_close(vdev_t *vd)
{
    uint64_t c;

    for (c = 0; c < vd->vdev_children; c++)
        vdev_close(vd->vdev_child[c]);
}

static void
vdev_mirror_child_done(zio_t *zio)
{
    mirror_map_t *mm = zio->io_private;

    mm->mm_error = zio->io_error;
    mm->mm_tried = 1;
    mm->mm_skipped = 0;
}

static void
vdev_mirror_scrub_done(zio_t *zio)
{
    mirror_map_t *mm = zio->io_private;

    if (zio->io_error == 0) {
        zio_t *pio = zio->io_parent;
        mutex_enter(&pio->io_lock);
        bcopy(zio->io_data, pio->io_data, pio->io_size);
        mutex_exit(&pio->io_lock);
    }

    zio_buf_free(zio->io_data, zio->io_size);

    mm->mm_error = zio->io_error;
    mm->mm_tried = 1;
    mm->mm_skipped = 0;
}

/*
 * Try to find a child whose DTL doesn't contain the block we want to read.
 * If we can't, try the read on any vdev we haven't already tried.
 */
static int
vdev_mirror_child_select(zio_t *zio)
{
    mirror_map_t *mm = zio->io_vsd;
    vdev_t *vd = zio->io_vd;
    vdev_t *cvd;
    uint64_t txg = zio->io_txg;
    int i, c;

    ASSERT(zio->io_bp == NULL || zio->io_bp->blk_birth == txg);

    /*
     * Select the child we'd like to read from absent any errors.
     * The current policy is to alternate sides at 8M granularity.
     * XXX -- investigate other policies for read distribution.
     */
    c = (zio->io_offset >> (SPA_MAXBLOCKSHIFT + 6)) % vd->vdev_children;

    /*
     * If this is a replacing vdev, always try child 0 (the source) first.
     */
    if (vd->vdev_ops == &vdev_replacing_ops)
        c = 0;

    /*
     * Try to find a child whose DTL doesn't contain the block to read.
     * If a child is known to be completely inaccessible (indicated by
     * vdev_is_dead() returning B_TRUE), don't even try.
     */
    for (i = 0; i < vd->vdev_children; i++, c++) {
        if (c >= vd->vdev_children)
            c = 0;
        if (mm[c].mm_tried || mm[c].mm_skipped)
            continue;
        cvd = vd->vdev_child[c];
        if (vdev_is_dead(cvd)) {
            mm[c].mm_error = ENXIO;
            mm[c].mm_tried = 1; /* don't even try */
            mm[c].mm_skipped = 1;
            continue;
        }
        if (!vdev_dtl_contains(&cvd->vdev_dtl_map, txg, 1))
            return (c);
        mm[c].mm_error = ESTALE;
        mm[c].mm_skipped = 1;
    }

    /*
     * Every device is either missing or has this txg in its DTL.
     * If we don't have any sibling replicas to consult, look for
     * any child we haven't already tried before giving up.
     */
    if (vd == vd->vdev_top || vd->vdev_parent->vdev_children <= 1) {
        for (c = 0; c < vd->vdev_children; c++) {
            if (!mm[c].mm_tried)
                return (c);
        }
    }

    /*
     * Every child failed.  There's no place left to look.
     */
    return (-1);
}

static void
vdev_mirror_io_start(zio_t *zio)
{
    vdev_t *vd = zio->io_vd;
    mirror_map_t *mm;
    int c, children;

    mm = vdev_mirror_map_alloc(zio);

    if (zio->io_type == ZIO_TYPE_READ) {
        if (zio->io_flags & ZIO_FLAG_SCRUB) {
            /*
             * For scrubbing reads we need to allocate a read
             * buffer for each child and issue reads to all
             * children.  If any child succeeds, it will copy its
             * data into zio->io_data in vdev_mirror_scrub_done.
             */
            for (c = 0; c < vd->vdev_children; c++) {
                zio_nowait(zio_vdev_child_io(zio, zio->io_bp,
                    vd->vdev_child[c], zio->io_offset,
                    zio_buf_alloc(zio->io_size), zio->io_size,
                    zio->io_type, zio->io_priority,
                    ZIO_FLAG_CANFAIL, vdev_mirror_scrub_done,
                    &mm[c]));
            }
            zio_wait_children_done(zio);
            return;
        }
        /*
         * For normal reads just pick one child.
         */
        c = vdev_mirror_child_select(zio);
        children = (c >= 0);
    } else {
        ASSERT(zio->io_type == ZIO_TYPE_WRITE);

        /*
         * If this is a resilvering I/O to a replacing vdev,
         * only the last child should be written -- unless the
         * first child happens to have a DTL entry here as well.
         * All other writes go to all children.
         */
        if ((zio->io_flags & ZIO_FLAG_RESILVER) &&
            vd->vdev_ops == &vdev_replacing_ops &&
            !vdev_dtl_contains(&vd->vdev_child[0]->vdev_dtl_map,
            zio->io_txg, 1)) {
            c = vd->vdev_children - 1;
            children = 1;
        } else {
            c = 0;
            children = vd->vdev_children;
        }
    }

    while (children--) {
        zio_nowait(zio_vdev_child_io(zio, zio->io_bp,
            vd->vdev_child[c], zio->io_offset, zio->io_data,
            zio->io_size, zio->io_type, zio->io_priority,
            ZIO_FLAG_CANFAIL, vdev_mirror_child_done, &mm[c]));
        c++;
    }

    zio_wait_children_done(zio);
}

static void
vdev_mirror_io_done(zio_t *zio)
{
    vdev_t *vd = zio->io_vd;
    vdev_t *cvd;
    mirror_map_t *mm = zio->io_vsd;
    int c;
    int good_copies = 0;
    int unexpected_errors = 0;

    ASSERT(mm != NULL);

    zio->io_error = 0;
    zio->io_numerrors = 0;

    for (c = 0; c < vd->vdev_children; c++) {
        if (mm[c].mm_tried && mm[c].mm_error == 0) {
            good_copies++;
            continue;
        }

        /*
         * We preserve any EIOs because those may be worth retrying;
         * whereas ECKSUM and ENXIO are more likely to be persistent.
         */
        if (mm[c].mm_error) {
            if (zio->io_error != EIO)
                zio->io_error = mm[c].mm_error;
            if (!mm[c].mm_skipped)
                unexpected_errors++;
            zio->io_numerrors++;
        }
    }

    if (zio->io_type == ZIO_TYPE_WRITE) {
        /*
         * XXX -- for now, treat partial writes as success.
         */
        /* XXPOLICY */
        if (good_copies != 0)
            zio->io_error = 0;
        ASSERT(mm != NULL);
        vdev_mirror_map_free(zio);
        zio_next_stage(zio);
        return;
    }

    ASSERT(zio->io_type == ZIO_TYPE_READ);

    /*
     * If we don't have a good copy yet, keep trying other children.
     */
    /* XXPOLICY */
    if (good_copies == 0 && (c = vdev_mirror_child_select(zio)) != -1) {
        ASSERT(c >= 0 && c < vd->vdev_children);
        cvd = vd->vdev_child[c];
        dprintf("%s: retrying i/o (err=%d) on child %s\n",
            vdev_description(zio->io_vd), zio->io_error,
            vdev_description(cvd));
        zio->io_error = 0;
        zio_vdev_io_redone(zio);
        zio_nowait(zio_vdev_child_io(zio, zio->io_bp, cvd,
            zio->io_offset, zio->io_data, zio->io_size,
            ZIO_TYPE_READ, zio->io_priority, ZIO_FLAG_CANFAIL,
            vdev_mirror_child_done, &mm[c]));
        zio_wait_children_done(zio);
        return;
    }

    /* XXPOLICY */
    if (good_copies)
        zio->io_error = 0;
    else
        ASSERT(zio->io_error != 0);

    if (good_copies && (spa_mode & FWRITE) &&
        (unexpected_errors || (zio->io_flags & ZIO_FLAG_RESILVER))) {
        /*
         * Use the good data we have in hand to repair damaged children.
         */
        for (c = 0; c < vd->vdev_children; c++) {
            /*
             * Don't rewrite known good children.
             * Not only is it unnecessary, it could
             * actually be harmful: if the system lost
             * power while rewriting the only good copy,
             * there would be no good copies left!
             */
            cvd = vd->vdev_child[c];

            if (mm[c].mm_error == 0) {
                if (mm[c].mm_tried)
                    continue;
                if (!vdev_dtl_contains(&cvd->vdev_dtl_map,
                    zio->io_txg, 1))
                    continue;
                mm[c].mm_error = ESTALE;
            }

            dprintf("%s resilvered %s @ 0x%llx error %d\n",
                vdev_description(vd),
                vdev_description(cvd),
                zio->io_offset, mm[c].mm_error);

            zio_nowait(zio_vdev_child_io(zio, zio->io_bp, cvd,
                zio->io_offset, zio->io_data, zio->io_size,
                ZIO_TYPE_WRITE, zio->io_priority,
                ZIO_FLAG_IO_REPAIR | ZIO_FLAG_CANFAIL |
                ZIO_FLAG_DONT_PROPAGATE, NULL, NULL));
        }
    }

    vdev_mirror_map_free(zio);
    zio_next_stage(zio);
}

static void
vdev_mirror_state_change(vdev_t *vd, int faulted, int degraded)
{
    if (faulted == vd->vdev_children)
        vdev_set_state(vd, VDEV_STATE_CANT_OPEN, VDEV_AUX_NO_REPLICAS);
    else if (degraded + faulted != 0)
        vdev_set_state(vd, VDEV_STATE_DEGRADED, VDEV_AUX_NONE);
    else
        vdev_set_state(vd, VDEV_STATE_HEALTHY, VDEV_AUX_NONE);
}

vdev_ops_t vdev_mirror_ops = {
    vdev_mirror_open,
    vdev_mirror_close,
    vdev_default_asize,
    vdev_mirror_io_start,
    vdev_mirror_io_done,
    vdev_mirror_state_change,
    VDEV_TYPE_MIRROR,   /* name of this vdev type */
    B_FALSE         /* not a leaf vdev */
};

vdev_ops_t vdev_replacing_ops = {
    vdev_mirror_open,
    vdev_mirror_close,
    vdev_default_asize,
    vdev_mirror_io_start,
    vdev_mirror_io_done,
    vdev_mirror_state_change,
    VDEV_TYPE_REPLACING,    /* name of this vdev type */
    B_FALSE         /* not a leaf vdev */
};