Closed yarikoptic closed 1 year ago
FWIW doing
/bin/ls -td *-*-* | while read d; do
( cd $d; gsha=$(git show-ref github/draft | awk '{print $1;}'); lsha=$(git show-ref heads/draft | awk '{print $1;}'); [ -n "$lsha" ] && [ "$lsha" == "$gsha" ] || { echo $d; datalad push --to=github; } );
done
now to go through all dandizarrs from newer to older and push whenever head is not the same on github
"interesting" is that it found quite a few to push. Lots (if not all) are from Dec 14th 2022, currently on
(dandisets-2) dandi@drogon:/mnt/backup/dandi/dandizarrs$ nl /tmp/dandizars.list | grep 2456b396-edb5-4344-adb1-5b05d1f32808
442 drwxr-sr-x 1 dandi dandi 134 Dec 14 03:29 2456b396-edb5-4344-adb1-5b05d1f32808
out of
(dandisets-2) dandi@drogon:/mnt/backup/dandi/dandizarrs$ nl /tmp/dandizars.list | tail -n 1
3939 drwxr-sr-x 1 dandi dandi 126 May 26 2022 1825d3a0-81f4-4fe3-923b-a6149d059601
so I did all the pushes etc, hopefully made it into a consistent state. Let's close for now
Run was going very well
```shell [INFO ] scanning for annexed files (this may take some time) [INFO ] Remote origin not usable by git-annex; setting annex-ignore [INFO ] https://github.com/dandizarrs/62f42eaa-495c-47dd-987a-d0c9f8e8919a/config download failed: Not Found [INFO ] access to 1 dataset sibling dandi-dandizarrs-dropbox not auto-enabled, enable with: | datalad siblings -d "/mnt/backup/dandi/dandisets/000108/sub-SChmi53/ses-20220927h13m20s15/micr/sub-SChmi53_ses-20220927h13m20s15_sample-SChmi53_25_stain-LEC_run-1_chunk-2_SPIM.ome.zarr" enable -s dandi-dandizarrs-dropbox install(ok): /mnt/backup/dandi/dandisets/000108/sub-SChmi53/ses-20220927h13m20s15/micr/sub-SChmi53_ses-20220927h13m20s15_sample-SChmi53_25_stain-LEC_run-1_chunk-2_SPIM.ome.zarr (dataset) Renaming remote references: 100% (4/4), done. Adding existing repo at 'sub-SChmi53/ses-20220927h13m20s15/micr/sub-SChmi53_ses-20220927h13m20s15_sample-SChmi53_25_stain-LEC_run-1_chunk-2_SPIM.ome.zarr' to the index add .gitmodules (non-large file; adding content to git repository) ok (recording state in git...) [WARNING] You appear to have cloned an empty repository. Clone attempt: 50%|█████████████████████████████████████████████████████████████████████████████████████████████ | 1.00/2.00 [00:00<00:00, 2.75 Candidate locations/s]2023-02-03T07:00:51-0500 [WARNING] datalad.gitrepo: You appear to have cloned an empty repository. [WARNING] Cloned /mnt/backup/dandi/dandisets/000108/sub-SChmi53/ses-20220927h13m20s15/micr/sub-SChmi53_ses-20220927h13m20s15_sample-SChmi53_25_stain-LEC_run-1_chunk-3_SPIM.ome.zarr but could not find a branch with commits 2023-02-03T07:00:51-0500 [WARNING] datalad.core.distributed.clone: Cloned /mnt/backup/dandi/dandisets/000108/sub-SChmi53/ses-20220927h13m20s15/micr/sub-SChmi53_ses-20220927h13m20s15_sample-SChmi53_25_stain-LEC_run-1_chunk-3_SPIM.ome.zarr but could not find a branch with commits install(ok): /mnt/backup/dandi/dandisets/000108/sub-SChmi53/ses-20220927h13m20s15/micr/sub-SChmi53_ses-20220927h13m20s15_sample-SChmi53_25_stain-LEC_run-1_chunk-3_SPIM.ome.zarr (dataset) fatal: 'sub-SChmi53/ses-20220927h13m20s15/micr/sub-SChmi53_ses-20220927h13m20s15_sample-SChmi53_25_stain-LEC_run-1_chunk-3_SPIM.ome.zarr' does not have a commit checked out 2023-02-03T07:00:52-0500 [ERROR ] backups2datalad: Job failed on input
https://github.com/dandizarrs/cfe2814c-005c-4b5d-85b5-7b1d316e027f is indeed empty. But the one we have locally is not:
so I guess we didn't push. The question why/how we didn't spot - probably happened in one of the crushed/killed cases... I will dig a little more but overall I think I would just ensure that we have everything pushed among present zarrs, and restart backup.