DF -A vs DF

[lucabisceglia]

Hi,

we need to migrate very old FAS with 7-mode through a tool like robocopy because 7MTT is not supported.

We received from customer an autosupport. 

My concern is about what to consider as space to migrate.

If I look in DF-A section I have bigger space Vs DF without any parameters.

What is the right space to consider ?

Many Thanks Luca

[paul_stejskal]

df -A = aggregate

df is just regular volume space.

ONTAP has two WAFL file systems at the aggregate and volume level (volumes within the aggregate).

So...both? Probably the total size of all volumes, but the aggregate can give you an idea too.

[lucabisceglia]

Hi,

the issue is that if I sum all the autosupport received from customer with:

DF-A total is 186TB

DF-S total is 106TB (used + deduplication)

There is a huge difference, I guess there's something hidden I don't know  at aggregate level.

Many thanks Luca 

[paul_stejskal]

df -s = df + sis (dedupe) space savings

[lucabisceglia]

Yes, my idea is to use DF-S instead of DF-A but I don't understand why there's such a difference.

Many thanks Luca

[paul_stejskal]

Do you have the AutoSupport ID by chance or serial number (if the system is sending AutoSupports)? It might just be easier to take a look.

[lucabisceglia]

Hi,

I noticed that all volumes have "fractional_reserve" set to 100%, is this the reason why I see such difference at aggregate level ?

Thanks Luca

[paul_stejskal]

That might be. That should thin provision the volumes if I remember correctly. Good find.

[lucabisceglia · Accepted solution]

Hi, probably I found the solution reading this KB:

https://mysupport.netapp.com/site/article?lang=en&page=%2FLegacy%2FONTAP%2F7Mode%2FDiscrepancies_between_aggregate_and_volume_when_using_the_df_comman...

All volumes of customer are space guarantee so DF total TB are equal to DF-A. 

I might say that for migration purposes the DF-S is the right choice.

Many thanks Luca