snappy vs. quicklz under fractal-tree layout

Description

Just did a test under fractal-tree layout as:
write/read nodes with:
nonleaf nums: 11 (each nonleaf size is ~ 4MB)
leaf nums: 80 (each leaf size is ~ 1MB)

using quicklz:

and using snappy:

the compressed datas size is same.
Snappy usually is faster than others, welcome the news and look forward to seeing it in tokudb.

-BohuTANG

Environment

None

Status

Assignee

Unassigned

Reporter

Bohu TANG

Labels

None

External issue ID

243

Freshdesk Tickets

None

Fix versions

Priority

Minor
Configure