Hello! I'm trying to download the workdir of a fai...
# nostupidquestions
a
Hello! I'm trying to download the workdir of a failed process from an s3 bucket, but it seems that all the files just contain a fusion link, e.g.
/fusion/s3/pixelgen-nf-work/work/bf/e3f3b69e0b382babd54a614fdb15f7/graph/PNA117_S01.report.json
. Is there a way to tell the aws cli to automatically resolve these links?
I ended up using a small script to convert the links:
Copy code
find $1 -type f | while read f; do
  content=$(cat "$f")
  if [[ "$content" == /fusion/s3/* ]]; then
    s3_uri=$(echo "$content" | sed 's|/fusion/s3/|s3://|')
    echo "Resolving $f -> $s3_uri"
    aws s3 cp "$s3_uri" "$f"
  fi
done
p
There have been discussions to have a Fusion CLI to be able to handle this kind of thing automatically. In fact it's possible that the
tw
CLI might already have a download function which handles Fusion files nicely? Not entirely sure.
100 rainbow 1
s
Phil, it'd be lovely if when you clicked on these files in Seqera Platform they automatically redirected to the actual file
100 rainbow 1
p
I know 100 rainbow I've been requesting that feature internally a bunch as well 😆
It's super irritating. I'm pretty sure it's on the the backlog somewhere already. Not sure if there's a feature request?
🤷 1
s
If there is one, I'll upvote it
p
I can't find any..
If you fancy making one, that'd be good. Sounds better coming from you than me 😅
s
Ugh, having to actually make a ticket
(feel free to make it clearer if it isn't)
p
hah, thank you 🙇🏻 🙏🏻
ok discussed and the work has now been scheduled, so hopefully should have something incoming in a few weeks / months
❤️ 1