Best way to migrate from file datastore to S3 + BinaryLess Replication setup?

[ryg457]

Dear community,

currently we have an AEM 6.5 deployment with the typical author, publisher instance seperation on AWS where each instance has their data store, segment store and node store on disk (EBS Volume).

The setup I want to move to is having the data store which takes up 95% of the crx-quickstart/repository directory migrated to S3 and then reduce the volume size significantly to achieve cost savings. The instances should also share this S3 bucket.

Our support team has set up the S3 connector as per documentation and migrated the data store, but seems to struggle with redeploying AEM on a fresh disk, installing necessary packages + code deployment and then reuse the state from the previous installation, but now with the data store offloaded.

I understand that it's easier to use this setup from the start, but there has to be a set by step approach to get to this point from on existing setup, right?

Would be very grateful for any high level steps on how to achieve this and remediate any potential issues or conflicts. How would we manage backup & restore in this scenario, as the backups are from the EBS disk but don't take into account the S3 bucket so things will get out of sync, right?

Regards,

ryg

[TarunKumar]

HI @ryg457 ,

Before doing migration of existing datastore to AWS S3, you might need to take a back up of the filedatastorage config.

Try to follow below steps:

1. Enable debug logs for the packages org.apache.jackrabbit.oak.blob  and org.apache.jackrabbit.oak.plugins.blob.

2. Shutdown AEM.

3. Back up the existing FDS configuration in …/crx-quickstart/install/.

4. Delete the FileDataStore.config in …/crx-quickstart/install/.

5. Copy the S3 connector to the …/crx-quickstart/install/folder.

6. Create the S3 configuration file into …/crx-quickstart/install/.
   
   For more details on it follow below link:
   https://experienceleague.adobe.com/en/docs/experience-cloud-kcs/kbarticles/ka-16057
   
   

-Tarun

[ryg457]

Hi Tarun,

unfortunately, your reply does not address my initial question at all. As mentioned we know how to use the S3 connector. The question is about migrating an existing setup to eventually free up the disk space saved through the migration and then also being able to manage all the operational tasks like backup and restore while using an external data store with binary less replication.

[kautuk_sahni]

@ryg457 

Did you find the suggestion helpful? Please let us know if you require more information. Otherwise, please mark the answer as correct for posterity. If you've discovered a solution yourself, we would appreciate it if you could share it with the community. Thank you!

[ryg457]

So far nobody has remotely answered my question unfortunately.