问题
Windows server 12r2 with python 2.7.10 and the aws cli tool installed. The following works:
aws s3 cp c:\a\a.txt s3://path/
I can upload that file without problem. What I want to do is upload a file from a mapped drive to an s3 bucket, so I tried this:
aws s3 cp s:\path\file s3://path/
and it works.
Now what I want to do and cannot figure out is how to not specify, but let it grab all file(s) so I can schedule this to upload the contents of a directory to my s3 bucket. I tried this:
aws s3 cp "s:\path\..\..\" s3://path/ --recursive --include "201512"
and I get this error "TOO FEW ARGUMENTS"
Nearest I can guess it's mad I'm not putting a specific file to send up, but I don't want to do that, I want to automate all things.
If someone could please shed some light on what I'm missing I would really appreciate it.
Thank you
回答1:
aws s3 cp "s:\path\..\..\" s3://path/ --recursive --include "201512"TOO FEW ARGUMENTS
This is because, in you command, double-quote(") is escaped with backslash(\), so local path(s:\path\..\..\) is not parsed correctly.
What you need to do is to escape backslash with double backslashes, i.e. :
aws s3 cp "s:\\path\\..\\..\\" s3://path/ --recursive --include "201512"
回答2:
Alternatively you can try 'mc' which comes as single binary is available for windows both 64bit and 32bit. 'mc' implements mirror, cp, resumable sessions, json parseable output and more - https://github.com/minio/mc
- 64-bit from https://dl.minio.io/client/mc/release/windows-amd64/mc.exe
- 32-bit from https://dl.minio.io/client/mc/release/windows-386/mc.exe
回答3:
In case this is useful for anyone else coming after me: Add some extra spaces between the source and target. I've been beating my head against running this command with every combination of single quotes, double quotes, slashes, etc:
aws s3 cp /home/<username>/folder/ s3://<bucketID>/<username>/archive/ --recursive --exclude "*" --include "*.csv"
And it would give me: "aws: error: too few arguments" Every. Single. Way. I. Tried.
So finally saw the --debug option in aws s3 cp help
so ran it again this way:
aws s3 cp /home/<username>/folder/ s3://<bucketID>/<username>/archive/ --recursive --exclude "*" --include "*.csv" --debug
And this was the relevant debug line:
MainThread - awscli.clidriver - DEBUG - Arguments entered to CLI: ['s3', 'cp', 'home/<username>/folder\xc2\xa0s3://<bucketID>/<username>/archive/', '--recursive', '--exclude', '*', '--include', '*.csv', '--debug']
I have no idea where \xc2\xa0 came from in between source and target, but there it is! Updated the line to add a couple extra spaces and now it runs without errors:
aws s3 cp /home/<username>/folder/ s3://<bucketID>/<username>/archive/ --recursive --exclude "*" --include "*.csv"
回答4:
Use aws s3 sync instead of aws s3 cp to copy the contents of a directory.
回答5:
I faced the same situation. Let share two scenarios I had tried to check the same code.
- Within bash please make sure you have AWS profile in place (use $aws configure). Also, make sure you use a proper proxy if applicable.
$aws s3 cp s3://bucket/directory/ /usr/home/folder/ --recursive --region us-east-1 --profile yaaagy
it worked.
Within a perl script
$cmd="$aws s3 cp s3://bucket/directory/ /usr/home/folder/ --recursive --region us-east-1 --profile yaaagy"
I enclosed it within "" and it was successful. Let me know if this works out for you.
来源：https://stackoverflow.com/questions/34535969/aws-cli-in-windows-wont-upload-file-to-s3-bucket