Lai, curious why this failedQA? I tested this with the dogfood server clone upgraded to 6.8.4 and did a delete and re-import. Was relatively quick and had no errors.
Sorry guys, this was a mistake made on my part when I was training someone. Changing this back.
Steps to retest: 1. Configure a Satellite with 60K entitled Candlepin consumers. 2. Double the default values for Tomcat's Java heap memory minimum and maximum: # grep JAVA_OPTS /etc/tomcat/tomcat.conf JAVA_OPTS="-Xms2048m -Xmx8192m" 3. Delete the manifest: # hammer subscription delete-manifest --organization-id 1 Expected Results: On a Satellite 6.8.2 with 60K Candlepin consumers, deleting the manifest takes a relatively long time, if it finishes successfully at all. On a Satellite 6.8.4 with 60K Candlepin consumers, manifest deletion should finish successfully in a timely fashion. Actual Results: On a Satellite 6.8.2 with 60K Candlepin consumers, deleting the manifest creates a task that runs for over 24 hours with no apparent progress. On a Satellite 6.8.4 snap 1 with 60K Candlepin consumers, deleting the manifest completes successfully in just under 2 hours. Note: In testing Bug #1919418, I found that deleting the manifest completes in about 10 seconds if the deletion is initiated immediately after refreshing the manifest. This behavior appears reproducible: deleting the manifest on a Satellite 6.8.4 with 60K entitled Candlepin consumers consistently takes about 10 seconds if the manifest is refreshed immediately before deleting, while deleting the manifest without refreshing it first consistently takes just under 2 hours. Verified on Satellite 6.8.4, snap 1 (candlepin 3.1.24-1).
Since the problem described in this bug report should be resolved in a recent advisory, it has been closed with a resolution of ERRATA. For information on the advisory (Satellite 6.8.4 Async Bug Fix Update), and where to find the updated files, follow the link below. If the solution does not work for you, open a new bug report. https://access.redhat.com/errata/RHBA-2021:0654