Good advice from Ananta.
You can save a lot of disk space by removing the stacks from batches you are no
longer going to use. These data are stored in the unique_tags table.
SELECT * FROM samples WHERE batch_id=X;
will give you a list of all sample IDs for a particular batch ID. Then, you can
take advantage of the sample IDs to delete rows from unique_tags.
First, check that you are getting the data you want:
SELECT * FROM unique_tags WHERE sample_id=Y LIMIT 10;
Now, do the deletion:
DELETE FROM unique_tags WHERE sample_id=Y;
This will take some time to complete. Repeat for each sample in the batch you
are removing. Of course, this will destroy data, so if you have the space, back
up the whole database first, then do your deletions before removing the back up.
There are lots of faster/fancier ways to do this if you want to learn more SQL.
Unfortunately there is no single button available to completely delete an
individual batch, if anyone has a script to do this, I'm sure it would be useful
to the group.
I tend to delete whole databases, which is a lot easier/faster. When you have
millions of reads/hundreds of samples, deleting all the records from the
database can take the SQL server a very long time. But, you can prune out a
bunch of data from the unique_tags data.
Best,
julian