why is HANA backup size smaller than before

[former_member612721]

Hello Community,

I come across a strange situation on HANA 2.0 DB

The tenant database backup was about 660 GB in size - which was an offline backup to backint.

After 8 hours or so (mostly with SAP system stopped), I ran a new backup and it generated a 650 GB size backup

Assuming that there was no data loss, is there any other reason why a backup could be smaller given these conditions?

Thank you,

Christian

[sumitjais · Accepted solution]

Hi Christian,

The backup that we see on the disk is just the summation of the allocated disk blocks (i.e output of

select sum(allocated_page_size) from m_converter_statistics + Other data) .

However, there are two things to be noted:

1. About other data -These blocks are not the only data that is backed up during a backup. A backup includes undo log too which keeps changing .
2. About the column data write operation - Write operations are done on less compressed delta storage (since main column storage is compressed and then merged to disk (called delta merge) after writing the data (of course in compressed way). This difference (i.e less compressed to compressed at the end) can result the variation of database size.

In your case,hope no log deletion job ran in the duration you mentioned.

But, the one or both the cases above can result the difference (based on amount of undo data and column data in delta mode) in database backup size.

It has nothing to do with data loss in the backup.

[kaus19d]

Hi alexandrunecula,

Well, I would say that your question is a valid one. To double sure what we can do is check the recover possibility details & other corresponding details or better try to trigger a 3rd backup & you can understand the scenario. You can check more details on HANA backup on SAP Note: 1642148

Thanks,

Kaushik