Migrating local storage to S3 - Self Hosted


#1

I am attempting to migrate my Razuna environment from local storage to S3. I have played around with Smart Folders and it does not accomplish what we need. I configured a private S3 bucket and configured the policy so only my IAM user has access. In the admin section of Razuna I have configured the necessary options for S3, and I have confirmed that new uploaded content is working fine.

My question is, can I migrate the local storage to the S3 bucket and have Razuna see the assets? Of course I have attempted to copy the content, reset the cache, but it does not seem to be that simple. Additionally I do not see anything in the logs. Has anyone else attempted this? Any assistance would be appreciated.

Thanks much


#2
  1. Activate the S3 storage in the backend admin and restart Razuna
  2. Upload your current local storage to the bucket (outside of Razuna)
  3. Once everything is correct, go to the DAM, then Administration and under
    the tab “Cloud maintenance” click on re-create all URLs

Hope this helps


#3

Thanks Nitai - I was able to follow your instructions, thanks for the help.

It looked like it found the videos and recreated the URLs, but each download and direct link do not work. I went through each tenant in the admin console and configured the cloud storage to point to the correct bucket. Again, it does not seem to work. One question I have is what kind of permissions need to be set on the bucket. If it is validated from the admin console are the permissions on the bucket set correctly? The error I am getting when I hit the direct links is NoSuchKeyThe specified key does not exist.. When I down load it seems to download, but the file size is always 1.2k and displays as corrupt in the player. Any ideas?


#4

You need to make sure to have the proper s3 credentials, region and also
the bucket name per tenant.


#5

I needed to upload the local storage in a different directory structure than was on the local storage. On the local storage the directory structure was associated with a specifc tenant (I am assuming) - i.e. it had /razunadata/2 /razunadata/3 etc… and then had specific GUIDs i.e. /razunadata/2FF8CA9DDB16B4BC7A24069080EE8B0E5 for the migration of the data I needed to upload just the GUIDs directly in the root of the S3 bucket. Then I used the Cloud Maintenance to recreate the URLs and that seemed to be successful.

We have started to go through testing in our test environment, and one thing I have noticed is that the indexing scheduled task is not longer working. I am not able to search for newly uploaded content.

I have tried to recreate the indexing with a new scheduled task, but that didn’t work. Any ideas here?

Thanks much for your help on this.