add apache/arrow repository
galsalomon66 opened this issue · comments
the apache/arrow (https://arrow.apache.org/install/) is an essential part of ceph/s3select project ( ceph/ceph#40802 ),it enables to read parquet objects (among other features)
parquet reader is a columnar reader, it enables to reduce IO significantly.
it also enables skipping data chunks according to data content.
it needs to add apache/arrow to the ceph-container process. (#1925)