Snapshot count

[sairajeev]

Why max snapshot count is 255 for flex volume ? is there any logical calculation on the count ?
I was asked in one of the interview - Couldn't give the answer.

[NetApp_SR]

With ONTAP 9.4 and later the Snapshot count has increased to 1023. It is a few less in some circumstances. The below KB may address some of  your questions.

Maximum number of snapshots supported by ONTAP
https://kb.netapp.com/Advice_and_Troubleshooting/Data_Storage_Software/ONTAP_OS/Maximum_number_of_snapshots_supported_by_ONTAP

[sairajeev]

Thanks

[paul_stejskal]

There were various internal WAFL and performance limitations, and it took a LOT of work to get it to the newer 1023 limit. What kind of interview are you talking about?

[sairajeev]

Interview was for Job switching - thanks for your reply 

[paul_stejskal]

You're welcome. Seems like an odd question to ask on an interview. That's a very specific internal only knowledge that only employees would in theory have access to.

[Ontapforrum · Accepted solution]

It's a very intriguing question and I would imagine this question be asked (IF AT ALL) to someone switching job internally at NetApp  Engg 🙂

Anyway, I am guessing there are some clues if you have read WAFL filesystem especially the ones that show up as patent in google search. There is a lot of information around how the meta-data file such as 'space map block entry' keeps tab on the block active/used/in-snapshot/free etc. One such information given in that google search talks about meta-data file called 'blkmap' file "block map file".

Considering (In theory): Before ontap 9.4, lets say Ontap/wafl used 8-bits to represent snapshots count, i.e 2^8 = 256 (n-1=255 snaps) and may be they used couple of more bits with Ontap 9.4, i.e 2^10 = 1024 (n-1=1023) snaps. Increasing snaps count must have been considered holistically b'cos it impacts number of internals. I don't know how much it makes sense but something to think about 🙂

[sairajeev]

Thanks  @Ontapforrum