Does Checkpoint Restart have any impact on index size?

Hi Everyone,

I saw a recent discussion that I thought others would be interested in.  A question was asked about Checkpoint Restart and does it have an impact on index sizes.  Below is a quick overview.  Let us know if you have any questions.

---

When looking at the documentation, there was a question on whether partial savesets are removed from the index when checkpoint restart is being used.

Does this mean that you should manually remove these partial savesets?

The answer in fact is no, since partial save sets contain valid data. You should not remove them manually as they will be deleting indexes for saved files. Aborted save sets are still removed from the index as before.

If the backup finished successfully, the sequence of partial save sets will constitute the original save set. The index database will contain index records for all the files. Browsing data will show all the files as if there are no partial save sets, but the original save set.

If the backup did not finish successfully, the index database will contain records for files already saved. These files may be browseable as well, or the files can be displayed using nsrinfo. Partial save set can always be recovered using save set recover.

To summarize, if you have unfinished sequences of partial save sets, you will have more index records for files that can now be recovered. If you have finished sequences of partial save sets, you will have as many index records as if your backup succeeded on the first try.

Browse and retention policy is applied to partial save sets in the same way as successful save sets.

Taking all this into consideration, there still is the time saving of not having to restart from the beginning.

----

Hope this helps.  Let us know if you have any questions.

Sherry