Hi Jim-

I just realized this didn't make it into our tree.  It's now in testing, 
and will get merged in the next window.  D'oh!

sage


On Fri, 30 Nov 2012, Jim Schutt wrote:

> Add libceph support for a new CRUSH tunable recently added to Ceph servers.
> 
> Consider the CRUSH rule
>   step chooseleaf firstn 0 type <node_type>
> 
> This rule means that <n> replicas will be chosen in a manner such that
> each chosen leaf's branch will contain a unique instance of <node_type>.
> 
> When an object is re-replicated after a leaf failure, if the CRUSH map uses
> a chooseleaf rule the remapped replica ends up under the <node_type> bucket
> that held the failed leaf.  This causes uneven data distribution across the
> storage cluster, to the point that when all the leaves but one fail under a
> particular <node_type> bucket, that remaining leaf holds all the data from
> its failed peers.
> 
> This behavior also limits the number of peers that can participate in the
> re-replication of the data held by the failed leaf, which increases the
> time required to re-replicate after a failure.
> 
> For a chooseleaf CRUSH rule, the tree descent has two steps: call them the
> inner and outer descents.
> 
> If the tree descent down to <node_type> is the outer descent, and the descent
> from <node_type> down to a leaf is the inner descent, the issue is that a
> down leaf is detected on the inner descent, so only the inner descent is
> retried.
> 
> In order to disperse re-replicated data as widely as possible across a
> storage cluster after a failure, we want to retry the outer descent. So,
> fix up crush_choose() to allow the inner descent to return immediately on
> choosing a failed leaf.  Wire this up as a new CRUSH tunable.
> 
> Note that after this change, for a chooseleaf rule, if the primary OSD
> in a placement group has failed, choosing a replacement may result in
> one of the other OSDs in the PG colliding with the new primary.  This
> requires that OSD's data for that PG to need moving as well.  This
> seems unavoidable but should be relatively rare.
> 
> Signed-off-by: Jim Schutt <jasc...@sandia.gov>
> ---
>  include/linux/ceph/ceph_features.h |    4 +++-
>  include/linux/crush/crush.h        |    2 ++
>  net/ceph/crush/mapper.c            |   13 ++++++++++---
>  net/ceph/osdmap.c                  |    6 ++++++
>  4 files changed, 21 insertions(+), 4 deletions(-)
> 
> diff --git a/include/linux/ceph/ceph_features.h 
> b/include/linux/ceph/ceph_features.h
> index dad579b..61e5af4 100644
> --- a/include/linux/ceph/ceph_features.h
> +++ b/include/linux/ceph/ceph_features.h
> @@ -14,13 +14,15 @@
>  #define CEPH_FEATURE_DIRLAYOUTHASH  (1<<7)
>  /* bits 8-17 defined by user-space; not supported yet here */
>  #define CEPH_FEATURE_CRUSH_TUNABLES (1<<18)
> +#define CEPH_FEATURE_CRUSH_TUNABLES2 (1<<25)
>  
>  /*
>   * Features supported.
>   */
>  #define CEPH_FEATURES_SUPPORTED_DEFAULT  \
>       (CEPH_FEATURE_NOSRCADDR |        \
> -      CEPH_FEATURE_CRUSH_TUNABLES)
> +      CEPH_FEATURE_CRUSH_TUNABLES |   \
> +      CEPH_FEATURE_CRUSH_TUNABLES2)
>  
>  #define CEPH_FEATURES_REQUIRED_DEFAULT   \
>       (CEPH_FEATURE_NOSRCADDR)
> diff --git a/include/linux/crush/crush.h b/include/linux/crush/crush.h
> index 25baa28..6a1101f 100644
> --- a/include/linux/crush/crush.h
> +++ b/include/linux/crush/crush.h
> @@ -162,6 +162,8 @@ struct crush_map {
>       __u32 choose_local_fallback_tries;
>       /* choose attempts before giving up */ 
>       __u32 choose_total_tries;
> +     /* attempt chooseleaf inner descent once; on failure retry outer 
> descent */
> +     __u32 chooseleaf_descend_once;
>  };
>  
>  
> diff --git a/net/ceph/crush/mapper.c b/net/ceph/crush/mapper.c
> index 35fce75..96c8a58 100644
> --- a/net/ceph/crush/mapper.c
> +++ b/net/ceph/crush/mapper.c
> @@ -287,6 +287,7 @@ static int is_out(const struct crush_map *map, const 
> __u32 *weight, int item, in
>   * @outpos: our position in that vector
>   * @firstn: true if choosing "first n" items, false if choosing "indep"
>   * @recurse_to_leaf: true if we want one device under each item of given type
> + * @descend_once: true if we should only try one descent before giving up
>   * @out2: second output vector for leaf items (if @recurse_to_leaf)
>   */
>  static int crush_choose(const struct crush_map *map,
> @@ -295,7 +296,7 @@ static int crush_choose(const struct crush_map *map,
>                       int x, int numrep, int type,
>                       int *out, int outpos,
>                       int firstn, int recurse_to_leaf,
> -                     int *out2)
> +                     int descend_once, int *out2)
>  {
>       int rep;
>       unsigned int ftotal, flocal;
> @@ -399,6 +400,7 @@ static int crush_choose(const struct crush_map *map,
>                                                        x, outpos+1, 0,
>                                                        out2, outpos,
>                                                        firstn, 0,
> +                                                      
> map->chooseleaf_descend_once,
>                                                        NULL) <= outpos)
>                                                       /* didn't get leaf */
>                                                       reject = 1;
> @@ -422,7 +424,10 @@ reject:
>                                       ftotal++;
>                                       flocal++;
>  
> -                                     if (collide && flocal <= 
> map->choose_local_tries)
> +                                     if (reject && descend_once)
> +                                             /* let outer call try again */
> +                                             skip_rep = 1;
> +                                     else if (collide && flocal <= 
> map->choose_local_tries)
>                                               /* retry locally a few times */
>                                               retry_bucket = 1;
>                                       else if 
> (map->choose_local_fallback_tries > 0 &&
> @@ -485,6 +490,7 @@ int crush_do_rule(const struct crush_map *map,
>       int i, j;
>       int numrep;
>       int firstn;
> +     const int descend_once = 0;
>  
>       if ((__u32)ruleno >= map->max_rules) {
>               dprintk(" bad ruleno %d\n", ruleno);
> @@ -544,7 +550,8 @@ int crush_do_rule(const struct crush_map *map,
>                                                     curstep->arg2,
>                                                     o+osize, j,
>                                                     firstn,
> -                                                   recurse_to_leaf, c+osize);
> +                                                   recurse_to_leaf,
> +                                                   descend_once, c+osize);
>                       }
>  
>                       if (recurse_to_leaf)
> diff --git a/net/ceph/osdmap.c b/net/ceph/osdmap.c
> index 3124b71..72f878e 100644
> --- a/net/ceph/osdmap.c
> +++ b/net/ceph/osdmap.c
> @@ -170,6 +170,7 @@ static struct crush_map *crush_decode(void *pbyval, void 
> *end)
>          c->choose_local_tries = 2;
>          c->choose_local_fallback_tries = 5;
>          c->choose_total_tries = 19;
> +     c->chooseleaf_descend_once = 0;
>  
>       ceph_decode_need(p, end, 4*sizeof(u32), bad);
>       magic = ceph_decode_32(p);
> @@ -336,6 +337,11 @@ static struct crush_map *crush_decode(void *pbyval, void 
> *end)
>          dout("crush decode tunable choose_total_tries = %d",
>               c->choose_total_tries);
>  
> +     ceph_decode_need(p, end, sizeof(u32), done);
> +     c->chooseleaf_descend_once = ceph_decode_32(p);
> +     dout("crush decode tunable chooseleaf_descend_once = %d",
> +          c->chooseleaf_descend_once);
> +
>  done:
>       dout("crush_decode success\n");
>       return c;
> -- 
> 1.7.8.2
> 
> 
> --
> To unsubscribe from this list: send the line "unsubscribe ceph-devel" in
> the body of a message to majord...@vger.kernel.org
> More majordomo info at  http://vger.kernel.org/majordomo-info.html
> 
> 
--
To unsubscribe from this list: send the line "unsubscribe ceph-devel" in
the body of a message to majord...@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

Reply via email to