[Migrating Clearcanvas to dcm4chee 2.x] Files and Instance Table takes the DB size to 120 GB and above.

[Rady]

Like to share my views on migrating from Clearcanvas to dcm4chee, In one of our old site we had ClearCanvas running for 10 years - now they gave decided to move to dcm4chee to use services like auto routing, WADO, delayed compression and S3 Sync. 

In clearcanvas the database size was not big, couple of GBs, but after long migration process we realized the database size is upwards of 120 GB. After further investigation we found only two files take a lot of storage space Files and Instance tables. each is above 40GB.  Because of such massive size my delayed compression throws Memory Issues and not working. 

An SQL statement,  takes more than 6 minutes. 

SELECT files.pk, files.created_time, files.md5_check_time, files.filepath, files.file_tsuid, files.file_md5, files.file_status, files.file_size, files.instance_fk, files.filesystem_fk
FROM files , fileSystem, instance
WHERE (files.file_status = 0
AND files.file_tsuid IN ('1.2.840.10008.1.2', '1.2.840.10008.1.2.1', '1.2.840.10008.1.2.2')
AND fileSystem.pk = 1 AND instance.sop_cuid = '1.2.840.10008.5.1.4.1.1.4' AND files.created_time < '2017-01-30 23:59:59'
AND files.filesystem_fk=fileSystem.pk AND files.instance_fk=instance.pk)

I would like to know if there options to 

1. Run delayed compression after such large database size. unfortunately all our data migrated are Raw and runs upwards of 12 TB of storage. Like to compress and store them in s3. 
   
2. Is there a way to avoid creating Files and Instance records, is there a need to keep track of image level in database - to me this looks like a potential problem given the fact current modalities generate 1000s of files per study. I have a Philips MRI generate 50,000 images / study for Functional Analysis. 
   

Love to hear experts opinion on this topic. 

with regards
Rady