jackye1995 commented on code in PR #7105:
URL: https://github.com/apache/iceberg/pull/7105#discussion_r1136024456


##########
format/partition-statistics-spec.md:
##########
@@ -0,0 +1,43 @@
+---
+title: "Partition Statistics Spec"
+url: partition-statistics-spec
+toc: true
+disableSidebar: true
+---
+<!--
+ - Licensed to the Apache Software Foundation (ASF) under one or more
+ - contributor license agreements.  See the NOTICE file distributed with
+ - this work for additional information regarding copyright ownership.
+ - The ASF licenses this file to You under the Apache License, Version 2.0
+ - (the "License"); you may not use this file except in compliance with
+ - the License.  You may obtain a copy of the License at
+ -
+ -   http://www.apache.org/licenses/LICENSE-2.0
+ -
+ - Unless required by applicable law or agreed to in writing, software
+ - distributed under the License is distributed on an "AS IS" BASIS,
+ - WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+ - See the License for the specific language governing permissions and
+ - limitations under the License.
+ -->
+
+# Partition Statistics file format
+
+This is a specification for partition statistics files. It is designed to 
store statistics information 
+for every partition value as a row in the **table default format** sorted 
based on the `partition`.
+
+The schema of the file is as follows:
+
+| Field Id | Field Name              | Field Type   | isOptional | Doc         
                                                                                
                                |
+|----------|-------------------------|--------------|------------|-----------------------------------------------------------------------------------------------------------------------------|
+| 1        | partition               | StructType   | false      | See 
[PartitionData](https://github.com/apache/iceberg/blob/master/core/src/main/java/org/apache/iceberg/PartitionData.java)
 |                                                                              
    
+| 2        | spec_id                 | IntegerType  | false      | partition 
spec id                                                                         
                                  |
+| 3        | data_record_count       | LongType     | false      | Count of 
records in data files                                                           
                                   |                                            
   
+| 4        | data_file_count         | IntegerType  | false      | Count of 
data files                                                                      
                                   |
+| 5        | pos_delete_record_count | LongType     | false      | Count of 
records in position delete files                                                
                                   |                                            
              
+| 6        | pos_delete_file_count   | IntegerType  | false      | Count of 
position delete files                                                           
                                   |                                   
+| 7        | eq_delete_record_count  | LongType     | false      | Count of 
records in equality delete files                                                
                                   |                                  
+| 8        | eq_delete_file_count    | IntegerType  | false      | Count of 
equality delete files                                                           
                                   |                                

Review Comment:
   how do we draw the line between what is optional and what is not? 
   
   I see your have the following phases in the design:
   
   > Phase2 implementation can be about adding more fields to the partition 
stats like latest sequence number for the partition, file_size_in_bytes and 
other metrics from the data_file spec 
(https://iceberg.apache.org/spec/#manifests) to support special use cases for 
the users.
   > 
   > Phase3 implementation can be for writing partition level puffin files and 
adding the puffin file location to the schema of partition stats. 
   
   Do we still plan to introduce puffin files to this? Based on the discussion 
in the doc, it seems to have scalability concerns which I agree. So if we want 
to add at least NDV and bloom filter, feels like those needs to be optional 
stats directly in the partition stats file?



-- 
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: [email protected]

For queries about this service, please contact Infrastructure at:
[email protected]


---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]

Reply via email to