System state rollback after modification failure
First Claim
Patent Images
1. A computer implemented method for rolling back a system state after a modification failure, the method comprising the steps of:
- a rollback manager creating a restore point on a computer;
the rollback manager storing a reboot indicator in non-revertible storage;
the rollback manager storing initial audit information in non-revertible storage, the initial audit information comprising an enumeration of at least some items present in the system state before deployment of a modification, the items not intended to be modified by the deployment;
the rollback manager monitoring the reboot indicator during deployment of the modification, the monitoring comprising;
the rollback manager detecting a reboot of the computer, the reboot having occurred during the deployment of the modification; and
the rollback manager determining based at least in part on the reboot indicator whether the reboot was expected or whether the reboot was unexpected; and
responsive at least in part to determining that no unexpected reboot occurred during the deployment of the modification;
the rollback manager re-auditing the computer, the re-auditing comprising deeming the computer stable based on comparing each of the enumerated items in the stored initial audit information with items present in the system state after deployment of the modification to determine missing items; and
responsive to the rollback manager not deeming the computer stable, the rollback manager rolling back the system state of the computer according to the restore point.
7 Assignments
0 Petitions
Accused Products
Abstract
A rollback manager creates a restore point on a computer. The rollback manager stores a reboot indicator in non-revertible storage. The rollback manager monitors the reboot indicator to detect an unexpected reboot during deployment of a modification. The rollback manager configures the computer responsive to the reboot indicator. The rollback manager can roll back the system state, responsive to detecting a failed modification deployment. The rollback manager can deem the computer to be stable, responsive to detecting a successful modification deployment.
-
Citations
21 Claims
-
1. A computer implemented method for rolling back a system state after a modification failure, the method comprising the steps of:
-
a rollback manager creating a restore point on a computer; the rollback manager storing a reboot indicator in non-revertible storage; the rollback manager storing initial audit information in non-revertible storage, the initial audit information comprising an enumeration of at least some items present in the system state before deployment of a modification, the items not intended to be modified by the deployment; the rollback manager monitoring the reboot indicator during deployment of the modification, the monitoring comprising; the rollback manager detecting a reboot of the computer, the reboot having occurred during the deployment of the modification; and the rollback manager determining based at least in part on the reboot indicator whether the reboot was expected or whether the reboot was unexpected; and responsive at least in part to determining that no unexpected reboot occurred during the deployment of the modification; the rollback manager re-auditing the computer, the re-auditing comprising deeming the computer stable based on comparing each of the enumerated items in the stored initial audit information with items present in the system state after deployment of the modification to determine missing items; and responsive to the rollback manager not deeming the computer stable, the rollback manager rolling back the system state of the computer according to the restore point. - View Dependent Claims (2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16)
-
-
17. A computer readable storage medium containing an executable computer program product for rolling back a system state after a modification failure, the computer program product comprising:
-
program code for creating a restore point on a computer; program code for storing a reboot indicator in non-revertible storage; program code for storing initial audit information in non-revertible storage, the initial audit information comprising an enumeration of at least some items present in the system state before deployment of a modification, the items not intended to be modified by the deployment; program code for monitoring the reboot indicator during deployment of the modification, the monitoring comprising; detecting a reboot of the computer, the reboot having occurred during the deployment of the modification; and determining based at least in part on the reboot indicator whether the reboot was expected or whether the reboot was unexpected; and program code for, responsive at least in part to determining that no unexpected reboot occurred during the deployment of the modification; re-auditing the computer, the re-auditing comprising deeming the computer stable based on comparing each of the enumerated items in the stored initial audit information with items present in the system state after deployment of the modification to determine missing items; and responsive to not deeming the computer stable, rolling back the system state of the computer according to the restore point. - View Dependent Claims (18, 19)
-
-
20. A computer system including a computer to support code execution for rolling back a system state after a modification failure, the computer system comprising:
-
a creation module, configured to create a restore point on a computer; a storage module, configured to store a reboot indicator in non-revertible storage, the storage module being communicatively coupled to the creation module; an auditing module, configured to store initial audit information in non-revertible storage, the initial audit information comprising an enumeration of at least some items present in the system state before deployment of a modification, the items not intended to be modified by the deployment; a monitoring module, configured to monitor the reboot indicator during deployment of the modification, further configured to detect a reboot of the computer, further configured to determine based at least in part on the reboot indicator whether the reboot was expected or whether the reboot was unexpected, the monitoring module being communicatively coupled to the storage module; a re-auditing module, configured to, responsive at least in part to the monitoring module determining that no unexpected reboot occurred during the deployment of the modification, re-audit the computer, the re-auditing comprising deeming the computer stable based on comparing each of the enumerated items in the stored initial audit information with items present in the system state after deployment of the modification to determine missing items; and a rollback module, configured to, responsive at least in part to the re-auditing module not deeming the computer stable, roll back the system state of the computer according to the restore point. - View Dependent Claims (21)
-
Specification