How to Check That Backups Ran Successfully

1. Introduction

There are several different ways to find out whether your backup was successful, depending on what exactly you want to check: more options are available if you want to check how your scheduled, rather than manual, backups went. In addition, there are further resources if you are IT staff and want to check on your department/college's backups.

2. HFS Portal

The HFS Backup & Archive Services Portal gives a great deal of detail and is the recommended method of checking your backup status. If you click on a node name you will be taken to the Node Details page for that node. Here there are a number of tabs that contain information about backups: 

  • Summary - here you can see when the last backup was completed, the status of your last scheduled backup, a summary of backups of individual filespaces, and how much newly-sent data is on disk waiting to be committed to tape.
  • Filespaces - this provides more detail on backups of individual filespaces including how much data is held and the time and duration of the last backup.
  • Activity - this shows all backup and restore sessions.
  • Backup and restore - this shows the same information as the activity tab but as a graph.
  • Scheduled - here you can see times and statuses of recent and near-future scheduled backups. This shows the time when the backup is scheduled to start, and whether schedules have been completed, missed, failed or whether there was another problem. Please note that a completed schedule does not mean that every single file was successfully backed up; the sections below list methods that enable you to check for specific files.

Please note that this data is only extracted from the servers periodically: it can be up to 10-15 minutes out of date (or an hour in the case of the backup and restore tab).

3. IT Services Self-Registration Pages

The IT Services Self-Registration Pages offer three different options for checking what the HFS is holding in your TSM account(s). These are good for a general check, but unlike the HFS Portal do not list backup sessions that failed to complete, nor data that has not yet been committed to tape.

  • Show client's recent activity - the last 30 days of activity, both backups and restores; however, sessions that were cut off (for example if an account reached the daily limit during a backup) are not listed.
  • Show client's scheduled backups - times and statuses of recent and near-future scheduled backups. This shows the time when the backup is scheduled to start, and whether schedules have been completed, missed, failed or whether there was another problem. Please note that a completed schedule does not mean that every single file was successfully backed up; the sections below list methods that enable you to check for specific files.
  • Show client's data held on tape - data sent to the HFS of which we have three copies on tape. Please note that this may not include the most recent backup, because data is initially sent to HFS disk and only committed to tape overnight.

4. Test a restore

You can see the details of exactly what we have for your machine by running TSM and clicking on Restore. Under 'File Level' you will be able to see your drive(s) and what files we have for them, without the need for restoring any data. See further the examples for "Restoring selected files and directories" for Windows, Mac, Linux and Solaris.

5. Schedule log

This method gives lots of details, but is only appropriate if you use the overnight schedules. During a scheduled backup, every file sent to the HFS by TSM is logged in the schedule log dsmsched.log, which is kept on your machine (see our list of TSM log files for how to find this file). Additionally, errors are logged to dsmerror.log, which you will find in the same place. If you open dsmsched.log in a text editor then you will be able to see exactly what was sent, and whether any files failed to get backed up.

Scroll to the end of dsmsched.log and you should see that the TSM scheduler has written summary information of the last scheduled backup similar to that below.

14-12-2010 23:03:04 --- SCHEDULEREC STATUS BEGIN
14-12-2010 23:03:04 Total number of objects inspected:   91,497
14-12-2010 23:03:04 Total number of objects backed up:      113
14-12-2010 23:03:04 Total number of objects updated:          0
14-12-2010 23:03:04 Total number of objects rebound:          0
14-12-2010 23:03:04 Total number of objects deleted:          0
14-12-2010 23:03:04 Total number of objects expired:         53
14-12-2010 23:03:04 Total number of objects failed:           6
14-12-2010 23:03:04 Total number of bytes transferred:    19.38 MB
14-12-2010 23:03:04 Data transfer time:                    1.54 sec
14-12-2010 23:03:04 Network data transfer rate:        12,821.52 KB/sec
14-12-2010 23:03:04 Aggregate data transfer rate:        114.39 KB/sec
14-12-2010 23:03:04 Objects compressed by:                    0%
14-12-2010 23:03:04 Elapsed processing time:           00:02:53
14-12-2010 23:03:04 --- SCHEDULEREC STATUS END
14-12-2010 23:03:04 --- SCHEDULEREC OBJECT END WEEKLY_23_00 14-12-2010 23:00:00
14-12-2010 23:03:04 Scheduled event 'WEEKLY_23_00' completed successfully.
14-12-2010 23:03:04 Sending results for scheduled event 'WEEKLY_23_00'.
14-12-2010 23:03:04 Results sent to server for scheduled event 'WEEKLY_23_00'.

Firstly, check whether TSM deemed the schedule to be completed or not. The third line from the bottom indicates that the schedule completed successfully. This is likely to mean that most of your files were backed up, but on the other hand it does not necessarily mean that all files were sent to the HFS.

It is important, therefore, to determine if any files have failed to be backed up. If the scheduled completed successfully but the number of objects listed under Total number of objects failed is greater than zero, then scroll up or search for the next occurrence of the word 'fail' to determine which file(s) is/are affected. The most common reason for a file failing to be backed up is that it is held open (locked) by another program. For example:

14-12-2010 23:02:59 ANS1228E Sending of object '\\oucs-jason\c$\Documents and Settings\jasonz\NTUSER.DAT' failed
14-12-2010 23:02:59 ANS4987E Error processing '\\oucs-jason\c$\Documents and Settings\jasonz\NTUSER.DAT': the object 
is in use by another process 

In some cases this is unavoidable: the file dsmerror.log should be ignored if it fails and should not be excluded from the backup. In other cases, the choice is either to close the problem program before the backup, or to exclude the problem file from backup. For how to do this, please see our page on how to exclude files, folders and drives from backup.

Other failures and errors will be accompanied by an error message and a corresponding entry in dsmerror.log. If your schedule did not complete successfully then please see our page on troubleshooting TSM scheduled backup failure reports. Further help with any errors may be also found via the HFS knowledgebase.

Note that the schedule log file does not grow endlessly, but is pruned of entries older than the setting of the SchedlogRetention option. By default this is set to 30 days.

6. Further resources for IT support staff

IT staff may check the status of their college/departmental TSM accounts in the HFS Backup & Archive Services Portal  or via ITSS Unit-Level Information in the Registration system.

ITSS may list all the TSM accounts that fall under their unit. Note that in the ITSS Unit-Level Information unit ownership is inferred from the final part of the nodename, whereas in the portal the unit ownership is recorded for each node so the two can differ (although at node registration they are the same).

The HFS Portal offers ITSS to see the same information about individual nodes as end users can see about their own nodes. In addition it also provides a number of reports, available in PDF or Excel formats, on the unit's nodes.

In ITSS Unit-Level Information you can view a full list of TSM clients, or alternatively you can choose to see only those without a current owner or only those that are stale (i.e. not backed up for 90 days). For each account, you may view client information for the node, and also the same three items listed and explained in the section above headed 3. IT Services Self-Registration Pages: recent activity, scheduled backups and data held on tape.

Additionally, the HFS can send out weekly or monthly mailings to IT Support Staff which detail scheduled backup failures of their unit's machines (this is similar to the Scheduled Backups report which is available in the HFS Portal) . If you would like this, please mail hfs@ox.ac.uk and let us know your department (which may have TSM nodes that have several different TSM nodename suffixes); and whether you would prefer weekly or monthly mails.

Written by IT Services. Latest revision 13 September 2017