You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Import failures sharing cause data_unavailable, rolled up because 10 or more per-dataset issues were open.
NEMAR's bucket has no object for some of the keys this dataset's tree names, so it was not published (ADR 0064 withdraws a dataset below 90% of its data keys). This does NOT yet say the content is gone upstream: run nemar admin fleet content-recovery <id> to find out, per key, whether any recorded source is still readable.
Every further dataset with this cause is appended as a comment below; this list is not updated. Read the comments for the full set.
Per-dataset issues resume automatically once open ones drop to 5 or fewer, and this issue is closed automatically at the same moment -- it is a pressure-relief document, not a tracking issue, and its content stays readable once closed. Any dataset here that fails again gets its own issue.
See nemarOrg/nemar-cli#967 and docs/import-failure-procedure.md for the triage procedure.
Resolved: on007327 (source ds007327, DOI 10.18112/openneuro.ds007327.v1.1.0) was missing one canonical zero-byte key. Content recovery restored it and S3 read-back verified 0 bytes. Fleet key-registration confirmed and pushed all 1,009/1,009 annexed keys, with no missing or false claims. Import workflow 38803497011 completed successfully and NEMAR now reports complete. This resolves only on007327; other entries in this rollup remain open.
Per-dataset follow-up for on008837 (checked 2026-10-15):
The import remains quarantined at finalize with one of 151 annexed keys missing: SHA256E-s63010416--27105f1e90c24b4b9547e42241ef54a88aa2215dd2765a84e123c2de0a9df74c.set (sub-002/eeg/sub-002_task-vrCollisionHazard_eeg.set, 63,010,416 bytes declared).
The read-only NEMAR content-recovery report found no matching-size source object (unrecoverable=1, recovered=0, failed=0): /Volumes/S1/datasets/nemar-recovery-on008837-mhy7J3/report.json.
Anonymous OpenNeuro S3 HEAD for s3://openneuro.org/ds008837/sub-002/eeg/sub-002_task-vrCollisionHazard_eeg.set returned 0 bytes with the empty-object ETag; the sibling sub-001 recording returned 64,271,944 bytes. Public version listing showed only the current 0-byte object. Source page: https://openneuro.org/datasets/ds008837/versions/v1.1.2.
No S3 repair or key-registration change was made; the declared content remains unavailable from the checked upstream object.
The rollup stays open for unresolved on008837. The earlier comment records on007327 as resolved; it is not a dataset-specific issue.
Activity
nemar-bot commentedon Oct 7, 2026
Also failed with this cause (2026-10-07T21:07:21.296Z): on007327 (ds007327).
Workflow run: https://github.com/nemarDatasets/.github/actions/runs/37685615407
nemarAdmin commentedon Oct 15, 2026
Resolved: on007327 (source ds007327, DOI 10.18112/openneuro.ds007327.v1.1.0) was missing one canonical zero-byte key. Content recovery restored it and S3 read-back verified 0 bytes. Fleet key-registration confirmed and pushed all 1,009/1,009 annexed keys, with no missing or false claims. Import workflow 38803497011 completed successfully and NEMAR now reports complete. This resolves only on007327; other entries in this rollup remain open.
nemarAdmin commentedon Oct 15, 2026
Per-dataset follow-up for on008837 (checked 2026-10-15):
SHA256E-s63010416--27105f1e90c24b4b9547e42241ef54a88aa2215dd2765a84e123c2de0a9df74c.set(sub-002/eeg/sub-002_task-vrCollisionHazard_eeg.set, 63,010,416 bytes declared).unrecoverable=1,recovered=0,failed=0):/Volumes/S1/datasets/nemar-recovery-on008837-mhy7J3/report.json.s3://openneuro.org/ds008837/sub-002/eeg/sub-002_task-vrCollisionHazard_eeg.setreturned 0 bytes with the empty-object ETag; the sibling sub-001 recording returned 64,271,944 bytes. Public version listing showed only the current 0-byte object. Source page: https://openneuro.org/datasets/ds008837/versions/v1.1.2.The rollup stays open for unresolved on008837. The earlier comment records on007327 as resolved; it is not a dataset-specific issue.