[jira] [Updated] (CLOUDSTACK-2689) Create volume after upgrade from snapshot taken in 2.2.14 deletes the snapshot physically from sec. storage.
[ https://issues.apache.org/jira/browse/CLOUDSTACK-2689?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ] Nitin Mehta updated CLOUDSTACK-2689: Fix Version/s: 4.2.0 Affects Version/s: 4.2.0 > Create volume after upgrade from snapshot taken in 2.2.14 deletes the > snapshot physically from sec. storage. > > > Key: CLOUDSTACK-2689 > URL: https://issues.apache.org/jira/browse/CLOUDSTACK-2689 > Project: CloudStack > Issue Type: Bug > Security Level: Public(Anyone can view this level - this is the > default.) >Affects Versions: 4.2.0 >Reporter: Nitin Mehta >Assignee: Nitin Mehta > Fix For: 4.2.0 > > > In pre 3.0 release for any install be it fresh or upgraded the swift_id is 0 > bcz the swift_id is long and not the Java object Long and so by default gets > set to 0 instead of NULL in the DB. Now the code assumes that the snapshot is > present in swift and can be deleted from sec. storage after any operation on > the snapshot by checking the condition snapshot.getSwiftId() != null.It > should have checked for snapshot.getSwiftId() != null && > snapshot.getSwiftId() != 0. > Why this not an issue in fresh installs for 3.0 series ? > I think in 3.0 we added all the code for swift. > There was also a code change for Action which made swift_id as JAVA Long and > so for fresh installs in 3.0 series swift_id was NULL in the DB by default > and hence these operations succeeded without deleting snapshot physically > from sec. storage. But for any upgraded version from 2.2.8 onwards (when > swift_id got first introduced) this would be 0 and since we check only for > SwiftId() != null we will delete the snapshot from sec. storage. -- This message is automatically generated by JIRA. If you think it was sent incorrectly, please contact your JIRA administrators For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Updated] (CLOUDSTACK-2689) Create volume after upgrade from snapshot taken in 2.2.14 deletes the snapshot physically from sec. storage.
[ https://issues.apache.org/jira/browse/CLOUDSTACK-2689?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ] Nitin Mehta updated CLOUDSTACK-2689: Assignee: Nitin Mehta > Create volume after upgrade from snapshot taken in 2.2.14 deletes the > snapshot physically from sec. storage. > > > Key: CLOUDSTACK-2689 > URL: https://issues.apache.org/jira/browse/CLOUDSTACK-2689 > Project: CloudStack > Issue Type: Bug > Security Level: Public(Anyone can view this level - this is the > default.) >Reporter: Nitin Mehta >Assignee: Nitin Mehta > > In pre 3.0 release for any install be it fresh or upgraded the swift_id is 0 > bcz the swift_id is long and not the Java object Long and so by default gets > set to 0 instead of NULL in the DB. Now the code assumes that the snapshot is > present in swift and can be deleted from sec. storage after any operation on > the snapshot by checking the condition snapshot.getSwiftId() != null.It > should have checked for snapshot.getSwiftId() != null && > snapshot.getSwiftId() != 0. > Why this not an issue in fresh installs for 3.0 series ? > I think in 3.0 we added all the code for swift. > There was also a code change for Action which made swift_id as JAVA Long and > so for fresh installs in 3.0 series swift_id was NULL in the DB by default > and hence these operations succeeded without deleting snapshot physically > from sec. storage. But for any upgraded version from 2.2.8 onwards (when > swift_id got first introduced) this would be 0 and since we check only for > SwiftId() != null we will delete the snapshot from sec. storage. -- This message is automatically generated by JIRA. If you think it was sent incorrectly, please contact your JIRA administrators For more information on JIRA, see: http://www.atlassian.com/software/jira
