[LU-6374] replay-single test_20b: after 44416 > before 6528 Created: 17/Mar/15 Updated: 21/Dec/16 Resolved: 21/Dec/16 |
|
| Status: | Resolved |
| Project: | Lustre |
| Component/s: | None |
| Affects Version/s: | Lustre 2.7.0, Lustre 2.8.0 |
| Fix Version/s: | Lustre 2.10.0 |
| Type: | Bug | Priority: | Minor |
| Reporter: | Maloo | Assignee: | Niu Yawei (Inactive) |
| Resolution: | Fixed | Votes: | 0 |
| Labels: | None | ||
| Severity: | 3 |
| Rank (Obsolete): | 9223372036854775807 |
| Description |
|
This issue was created by maloo for Bob Glossman <bob.glossman@intel.com> This issue relates to the following test suite run: https://testing.hpdd.intel.com/test_sets/77b52408-ccc1-11e4-a8ca-5254006e85c2. This looks similar to The sub-test test_20b failed with the following error: after 44416 > before 6528 Please provide additional information about the failure here. Info required for matching: replay-single 20b |
| Comments |
| Comment by John Hammond [ 08/Jul/15 ] |
|
Another on 2.7.55+ https://testing.hpdd.intel.com/test_sets/4323393a-250b-11e5-8427-5254006e85c2. |
| Comment by Andreas Dilger [ 13/Aug/15 ] |
|
Also https://testing.hpdd.intel.com/test_sets/08cba93c-3db3-11e5-9e7f-5254006e85c2 on master. |
| Comment by James Nunez (Inactive) [ 10/Dec/15 ] |
|
More instances on master, all ZFS: |
| Comment by Jian Yu [ 25/Dec/15 ] |
|
More instance on master branch: |
| Comment by Niu Yawei (Inactive) [ 10/Nov/16 ] |
|
Hit on master: https://testing.hpdd.intel.com/test_sets/ed824568-a662-11e6-a6e7-5254006e85c2 |
| Comment by Niu Yawei (Inactive) [ 10/Nov/16 ] |
|
For zfs, we need wait for commit to release space, but looks wait_delete_completed_mds() didn't wait at all: wait_delete_completed_mds() {
local MAX_WAIT=${1:-20}
# for ZFS, waiting more time for DMUs to be committed
local ZFS_WAIT=${2:-5}
local mds2sync=""
local stime=$(date +%s)
local etime
local node
local changes
# find MDS with pending deletions
for node in $(mdts_nodes); do
changes=$(do_node $node "$LCTL get_param -n osc.*MDT*.sync_*" \
2>/dev/null | calc_sum)
if [[ $changes -eq 0 ]]; then
continue
fi
mds2sync="$mds2sync $node"
done
if [ -z "$mds2sync" ]; then
return <------------- before this return, we need to wait for zfs commit
fi
|
| Comment by Gerrit Updater [ 10/Nov/16 ] |
|
Niu Yawei (yawei.niu@intel.com) uploaded a new patch: http://review.whamcloud.com/23688 |
| Comment by Niu Yawei (Inactive) [ 17/Nov/16 ] |
|
The failure can be easily reproduced locally with zfs backend, and with above patch applied, I can't reproduce it anymore. |
| Comment by Gerrit Updater [ 19/Dec/16 ] |
|
Oleg Drokin (oleg.drokin@intel.com) merged in patch https://review.whamcloud.com/23688/ |
| Comment by Minh Diep [ 21/Dec/16 ] |
|
Landed in Lustre 2.10.0 |