1

I have a post-receive shell script as such:

#!/bin/bash

TGZNAME=imprv`date +%s`.tar.gz
TGZPATH=/var/git_out/imprv/
TGZFULL=$TGZPATH$TGZNAME
S3BUCKET=my-s3-bucket
S3PATH=s3://$S3BUCKET/imprv/

git --work-tree=/var/git_out/imprv/imprv --git-dir=/path/to/repositories/imprv/imprv.git checkout -f

echo " - creating tar "$TGZNAME >> /tmp/post-receive.log
tar -cjf $TGZFULL -C /var/git_out/imprv/imprv .
echo " - tar.gz created "$TGZFULL >> /tmp/post-receive.log

echo " - Uploading "$TGZFULL" to bucket "$S3PATH$TGZNAME >> /tmp/post-receive.log
s3cmd put $TGZFULL $S3PATH$TGZNAME --mime-type="application/gzip" >> /tmp/post-receive.log
echo " - operation complete" >> /tmp/post-receive.log

Nothing ends up in my bucket, but the tar file is created, and if I manually enter the s3cmd put command it works fine.

I am tailing the log file, and everything seems to proceed just fine, it sits and waits for the compressed file to be created before moving off to the next echo statement, so it's not trying to put the file before the tar.gz file is created. Thanks in advance.

Bernard Lechler
  • 277
  • 6
  • 19

1 Answers1

2

To upload file to s3 you should:

Configure CLI by running command aws configure then aws s3 sync <local_from> s3://<bucket_name> to sync local dir with your bucket.

UPDATE: Do not forget to install the required libraries: aws cli aws s3cmd

D.Dimitrioglo
  • 3,413
  • 2
  • 21
  • 41
  • Than you @D.Dimitrioglo. Your solution got me to thinking about how I had not configured the AWS CLI for the git user. Once that was done, it worked swimmingly. – Bernard Lechler May 27 '16 at 20:49