Skip to content

DAOS-18487 placement: fix a bug in shard remapping chain #17536

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Open
liuxuezhao wants to merge 1 commit into
base: release/2.6
Choose a base branch
from
+19 −2
Open

DAOS-18487 placement: fix a bug in shard remapping chain #17536

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
4 changes: 2 additions & 2 deletions src/object/obj_layout.c
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -115,8 +115,8 @@ obj_dump_grp_layout(daos_handle_t oh, uint32_t shard)
}

grp_idx = shard / obj->cob_grp_size;
D_INFO(DF_OID " shard %d, grp_idx %d, grp_size %d", DP_OID(obj->cob_md.omd_id), shard,
grp_idx, obj->cob_grp_size);
D_INFO(DF_OID " shard %d, grp_idx %d, grp_size %d, map_ver %d", DP_OID(obj->cob_md.omd_id),
shard, grp_idx, obj->cob_grp_size, obj->cob_version);
for (i = grp_idx * obj->cob_grp_size, nr = 0; nr < obj->cob_grp_size; i++, nr++) {
obj_shard = &obj->cob_shards->do_shards[i];
D_INFO("shard %d/%d/%d, tgt_id %d, rank %d, tgt_idx %d, "
Expand Down
17 changes: 17 additions & 0 deletions src/placement/pl_map_common.c
View file
Open in desktop
Original file line number Diff line number Diff line change
@@ -1,5 +1,6 @@
/**
* (C) Copyright 2016-2024 Intel Corporation.
* (C) Copyright 2026 Hewlett Packard Enterprise Development LP
*
* SPDX-License-Identifier: BSD-2-Clause-Patent
*/
Expand Down Expand Up @@ -378,6 +379,22 @@ determine_valid_spares(struct pool_target *spare_tgt, struct daos_obj_md *md,
f_shard->fs_fseq = spare_tgt->ta_comp.co_fseq;
f_shard->fs_status = spare_tgt->ta_comp.co_status;

/* For the case of 2nd remap, if the spare target is DOWN2UP need to set
* fs_down2up flag, to make it be able to set shard's po_rebuilding flag
* at the end.
* One example case -
* Target A is DOWN, rebuild completed and status changed to DOWNOUT
* Target B is DOWN, rebuild started but not completed but admin do the reint,
* its status change to UP and with DOWN2UP flag.
*
* In object layout calculation, one shard firstly located in Target A, but 1st
* remap to Target B, but still need to do 2nd remap. In this case should set
* fs_down2up flag which is not set in the 1st remap, to avoid not be able to set
* shard's po_rebuilding flag so will cause read from it (invalid place).
*/
if (pool_target_is_down2up(spare_tgt))
f_shard->fs_down2up = 1;

d_list_del_init(&f_shard->fs_list);
remap_add_one(remap_list, f_shard);

Expand Down
Loading