emkornfield commented on code in PR #11238:
URL: https://github.com/apache/iceberg/pull/11238#discussion_r1801477641


##########
format/puffin-spec.md:
##########
@@ -123,6 +123,49 @@ The blob metadata for this blob may include following 
properties:
 
 - `ndv`: estimate of number of distinct values, derived from the sketch.
 
+#### `delete-vector-v1` blob type
+
+A serialized delete vector (bitmap) that represents the positions of rows in a
+file that are deleted.  A set bit at position P indicates that the row at
+position P is deleted.
+
+The vector supports positive 64-bit positions, but is optimized for cases where
+most positions fit in 32 bits by using a collection of 32-bit Roaring bitmaps.
+64-bit positions are divided into a 32-bit "key" using the most significant 4
+bytes and a 32-bit sub-position using the least significant 4 bytes. For each
+key in the set of positions, a 32-bit Roaring bitmap is maintained to store a
+set of 32-bit sub-positions for that key.
+
+To test whether a certain position is set, its most significant 4 bytes (the
+key) are used to find a 32-bit bitmap and the least significant 4 bytes (the
+sub-position) are tested for inclusion in the bitmap. If a bitmap is not found
+for the key, then it is not set.
+
+The serialized blob contains:
+* The length of the vector and magic bytes stored as 4 bytes, little-endian

Review Comment:
   > The main use case for maintaining compatibility with the Delta format is 
that it's hard to move old jobs to new code. We see a similar issue in Hive to 
Iceberg migrations, too, where unknown older readers prevent migration because 
they are hard to track down and often read files directly from the backing 
object store. I'd like to avoid the same problem here, where all readers need 
to be identified and migrated at the same time. Doing that requires temporarily 
producing Delta metadata, which this makes possible.
   
   This will already be the case for an upgrade to Iceberg V3, where as 
currently proposed all position delete files will be migrated to DV?  Is it 
technically harder to do a similar migration of Delta lake to a new version if 
users would want to interop with Iceberg?   



-- 
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.

To unsubscribe, e-mail: issues-unsubscr...@iceberg.apache.org

For queries about this service, please contact Infrastructure at:
us...@infra.apache.org


---------------------------------------------------------------------
To unsubscribe, e-mail: issues-unsubscr...@iceberg.apache.org
For additional commands, e-mail: issues-h...@iceberg.apache.org

Reply via email to