You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
This would be fairly easy to do and would help for environments that generate and delete a large number of volumes and images. This has been tested in a production Newton environment with these additions to the current script:
DELETE FROM cinder.volume_admin_metadata where volume_id IN ( SELECT id from cinder.volumes WHERE deleted > 0 );
DELETE FROM cinder.volume_attachment where volume_id IN ( SELECT id from cinder.volumes WHERE deleted > 0 );
DELETE FROM cinder.volume_metadata where volume_id IN ( SELECT id from cinder.volumes WHERE deleted > 0 );
DELETE FROM cinder.volume_glance_metadata where snapshot_id IN ( SELECT id FROM cinder.snapshots where volume_id IN ( SELECT id from cinder.volumes WHERE deleted > 0 ));
DELETE FROM cinder.snapshot_metadata where snapshot_id IN ( SELECT id FROM cinder.snapshots where volume_id IN ( SELECT id from cinder.volumes WHERE deleted > 0 ));
DELETE FROM cinder.snapshots where volume_id IN ( SELECT id from cinder.volumes WHERE deleted > 0 );
DELETE FROM cinder.transfers where volume_id IN ( SELECT id from cinder.volumes WHERE deleted > 0 );
DELETE FROM cinder.volumes where deleted > 0;
DELETE FROM glance.image_locations where image_id IN ( SELECT id from glance.images WHERE deleted > 0 );
DELETE FROM glance.image_properties where image_id IN ( SELECT id from glance.images WHERE deleted > 0 );
DELETE FROM glance.image_tags where image_id IN ( SELECT id from glance.images WHERE deleted > 0 );
DELETE FROM glance.image_members where image_id IN ( SELECT id from glance.images WHERE deleted > 0 );
DELETE FROM glance.image_locations where image_id IN ( SELECT id from glance.images WHERE deleted > 0 );
DELETE FROM glance.images where deleted > 0;
If it's too confusing to have nova-instance-cleanup.sh do more than just clean up instances then these could be broken out in separate scripts, or the name of the script could be changed.
The text was updated successfully, but these errors were encountered:
Yeah we should look into using cinder-manage and other commands primarily as the data structure changes a lot in cinder. Let's think about this a little more and investigate how make these cleanup scripts part of the product via embedded archival
This would be fairly easy to do and would help for environments that generate and delete a large number of volumes and images. This has been tested in a production Newton environment with these additions to the current script:
If it's too confusing to have
nova-instance-cleanup.sh
do more than just clean up instances then these could be broken out in separate scripts, or the name of the script could be changed.The text was updated successfully, but these errors were encountered: