0

I'm trying to mount a runabove object storage, witch is based on openstack-swift, into my file system, thanks to S3QL.

I'm using Ubuntu server 14.04 LTS.

So I've mounted my object storage with:

$ mount.s3ql --authfile /home/sylvain/.s3ql.creds --log syslog swiftks://auth.runabove.io/SBG-1:data /home/sylvain/cloud
Using 2 upload threads.
Autodetected 4058 file descriptors available for cache entries
Using cached metadata.
Setting cache size to 233331 MB
Mounting filesystem...

It seems to runs without errors, but then when I try to list the content of my newly mounted endpoint: 

$ ls -la /home/sylvain/cloud/ 
total 0 
drwx------ 1 sylvain sylvain 0 juil.  6 10:00 lost+found

Nothing at all ... instead the fact that my container is NOT empty:

---data---
Object Count: 3140
Size: : 11,2 Go
Access: Private

Then I create a test file and unmount my endpoint:

$ echo "this is a test" > /home/sylvain/cloud/test.txt
$ umount.s3ql /home/sylvain/cloud

The newly created file does not appear into the horizon web interface... but if I re-mount the S3QL endpoint again, the file re-apears with its contents.

Any idea of what I've missed?

SylvainB
  • 135
  • 7

1 Answers1

0

S3QL is not a program to browse your existing objects, it's a file system that uses your container to store data. In other words, the only data that you can access with S3QL is the data that you have written into the S3QL file system, and the only way to read data that you've written using S3QL is to use S3QL.

Nikratio
  • 645
  • 5
  • 13
  • This is what I had deduced from the behavior, but thanks for the clarification anyway. The doc is not very obvious about this. – SylvainB Mar 16 '16 at 17:26