University Systems help centre
TSM Backup - Stale node notification process
A new process has been established to notify users whose computers are not contacting the TSM server to perform backups.
There have been a number of cases where users that thought their computer was backing up every night, when in fact they were not. University Systems is unable to identify all backup failures from the backup server, however, we can determine if a computer has logged in to the backup account. This information is used to identify computers that are completely failing to perform backups.
Types of errors that are caught are:
- Mismatched password between the backup node and the client backup software installation.
- Incorrectly configured client back up software.
- Client backup software is not running.
- Client system has been decommissioned and will not backup again.
See example below for a stale node that has never been backed up:
A TSM backup node will have the following life cycle:
- Active - node has connected to the backup server within the last 7 days.
- Stale - a node has not connected to the backup server for more than 7 days. Alerts are sent daily
- Acknowledged Stale - a node has not connected to the backup server for more than 7 days. No alerts are sent.
- Pending Deletion - a node has not connected to the backup server for more than 6 months. Alerts are sent daily informing users that the node is scheduled for deletion.
- Acknowledged Pending Deletion - a node has not connected to the backup server for more than 6 months. Alerts are not sent. A node automatically returns to Pending Deletion for 1 week prior to the deletion and cannot be acknowledged by the user.
- Deleting - A node will be deleted by an automated process as resources permit. No alerts are sent.
- Archived - Only archive data remains for a node. Archive data is self-expiring and when all of the archive data is removed, the empty node will be deleted. No alerts are sent.
Users will receive a minimum of:
- 1 alert for stale state
- 15 alerts for pending state
- 1 alert for the node being deleted
If the node is considered stale, the user will be sent an email notification with a link to log in to the interface using their NetLink ID and password.
The following description will be given:
"The backup node is considered stale when it fails to connect to the backup server for more than a week. After being stale for 4 months, the node will transition into a pending delete state for a time period and finally be deleted. Notifications will be sent every day warning of this process."
If the node is currently stale state, an opportunity to acknowledge the stale state is available by clicking on an Acknowledge button which indicates: “By acknowledging a stale alert you understand that no more notifications will be produced for the stale state.”
The current state will change to: Acknowledged
If the node is currently in the pending state, you can click on an Acknowledge button that indicates: “By acknowledging a pending alert you understand that no more notifications will be produced prior to the backup for this node being deleted.”
NOTE: A pending state can NOT be acknowledged during the two weeks prior to the node being deleted.
An owner can also request that a node be deleted by clicking on I understand and then Delete [Node Name].
The Current state will then be changed to Owner requested deletion. This will result in a final e-mail stating that the node will be deleting that evening.
