an old topic:why there is a size on the first snapshot?

[heightsnj]

After i took first snapshot on several volumes, then do "snap list...", some of snapshots have a few GB, some a few KB, MB, and some of them have 0 in size.

Why is that?

based on my understanding, the first snapshot would be all pointers to the original location of data, the size should be very small, or should be about the same on all first snapshots.

Where are these difference sizes coming from?

Can soembody please help me out here?

Thanks!

[aborzenkov]

Snapshot "size" is the amount of data that was changed (i.e. overwritten) or deleted since snapshot had been taken. Usually the older is snapshot, the more data was changed so it is shown as "bigger".

[heightsnj]

Yes, understood every single word you said here.

the size I am talking about is the one immediately shown after I just took the snapshot, there is no any change/deletion yet. they are supposed to be including no data, only pointers. Where these sizes / data coming from?

[J_curl]

compression,dedupe,read reallocate, any of those happening?  Will cause blocks to get charged to snapshot space.

[heightsnj]

NO, none of theose happening.

As I said, the size immediately shown right after I took snapshots on a number of volumes. there are no chances any of these would happen.

[naren_s]

Even the initial pointers occupy few KB, especially when file count is high.

[heightsnj]

No, it is not a few KB,  as I said in my previous posts,  it is quite some GB/MB/KB/0

[aborzenkov]

---

No, it is not a few KB,  as I said in my previous posts

---

If you show us what you see may be someone could answer your question.