Backup Estimation


#1

Hi, I want to perform a backup through asbackup but at first I want to estimate the volume of the backup file. On one of our production clusters we have 23 nodes with replication factor of 2, on the AMC I see 13.14 TB Cluster Disk Usage and 1.34TB Cluster Ram Usage. We use a single single namespace that has over 11 bilion of records. When I run the following command: asbackup -n <namespace_name> --estimate and I see to following output:

2017-02-02 12:47:53 GMT [INF] [31289] Backed up 10000 record(s), 0 secondary index(es), 0 UDF file(s) from 23 node(s), 6500722 byte(s) in total (~650 B/rec) 2017-02-02 12:47:53 GMT [INF] [31271] Estimated overall record size is 679 byte(s)

I don`t think that these numbers add up, how can I find an estimation for the backup that would be as close as possible to the real size?

EDIT: If I understood it correctly, the backup size is 679*11,000,000,000 (~number of records), which ends up as 7.469TB. Did I got it right?

Thanks, Alex.


#2

See http://www.aerospike.com/docs/tools/backup/asbackup.html

When passing the --estimate command line option to asbackup (and skipping --directory and --output-file), asbackup creates a temporary test backup of 10,000 records from the namespace. It then outputs, based on the observed record sizes, an estimate of the average size of a record in the backup. In order to estimate the total size of the backup file or files, multiply this size by the number of records in the namespace and add 10% for indexes and overhead.