Fragment size in SLP

[RyanLapuz]

*** moved to new post from https://vox.veritas.com/t5/NetBackup/Fragment-Size/m-p/839657 ***

Hi Marriane, I have a follow up question about fragment size and on your excellent explanation on the older post.

How is the fragment size being treated in a SLP on the below setup?
SLP1:
1st - backup to Disk (default fragment size of 500GB)
2nd - duplicate to LTO5tape (default fragment size of 1TB)

From your explanation, backing up a 1TB data with fragment size set to 500GB on the first instance, the backup data will be stored on 2 fragment with 500GB each.

For the 2nd instance (duplication) will it read that 2 fragments (as backup data) and store it as just 1 fragment? Or will it read the data just like how the 1st instance did?

[Marianne]

I don't think SLP duplications use STU fragment size.

SLPs breaks up the image into 'batches' for duplication. 

See: https://www.veritas.com/support/en_US/article.v84182756_nbu_core_admin_web

[Mike_Gavrilov · Accepted solution]

@Marianne It isn't completely true. NBU will gather small images into a batch till it's reached MAX_GB_SIZE_PER_DUPLICATION_JOB value, but if a single image is lager than MAX_GB_SIZE_PER_DUPLICATION_JOB NBU wouldn't split it:

MAX_GB_SIZE_PER_DUPLICATION_JOB 

This entry controls the maximum size of a duplication job. (When a single image is larger than the

maximum size, that one image will be put into its own duplication job.)

[Nicolai]

Its easy to see for a completed SLP, just use bpimagelis.This will show the copies and fragments sizes. 

This TN is handy for decoding output: http://www.veritas.com/docs/000032604