Optimize internal representation of snapshots

[redlicha]

Listing snapshot names is inefficient and hence takes long with large numbers of snapshots.

[wimpers]

@redlicha might this be related to openvstorage/framework#1747

Can you estimate how long this would take to fix?

[redlicha]

It might be related, depending on the number of snapshots present.
Currently the code

• gets a list of all snapshots by number
• sorts this list
• looks up the snapshot name by number
  .

Step 1 and 3 could be merged and depending on the actual implementation step 2 might be superfluous - +/- 2d.
This would however only improve list_snapshots. It might be desirable to change the internal representation from lists to an associative container to speed up lookups required by other calls as well, which requires code review (+/- 1d) + redesign (depends on the outcome of the code review).