Re: Review Request: HIVE-2246: Dedupe tables' column schemas from partitions in the metastore db
--- This is an automatically generated e-mail. To reply, visit: https://reviews.apache.org/r/1183/ --- (Updated 2011-08-08 20:55:11.546253) Review request for hive, Ning Zhang and Paul Yang. Changes --- added derby upgrade and revert-the-upgrade script Summary --- This patch tries to make minimal changes to the API while keeping migration short and somewhat easy to revert. The new schema can be described as follows: - CDS is a table corresponding to Column Descriptor objects. Currently, it only stores a CD_ID. - COLUMNS_V2 is a table corresponding to MFieldSchema objects, or columns. A Column Descriptor holds a list of columns. COLUMNS_V2 has a foreign key to the CD_ID to which it belongs. - SDS was modified to reference a Column Descriptor. So SDS now has a foreign key to a CD_ID which describes its columns. During migration, we create Column Descriptors for tables in a straightforward manner: their columns are now just wrapped inside a column descriptor. The SDS of partitions use their parent table's column descriptor, since currently a partition and its table share the same list of columns. When altering or adding a partition, give it it's parent table's column descriptor IF the columns they describe are the same. Otherwise, create a new column descriptor for its columns. When adding or altering a table, create a new column descriptor every time. Whenever you drop a storage descriptor (e.g, when dropping tables or partitions), check to see if the related column descriptor has any other references in the table. That is, check to see if any other storage descriptors point to that column descriptor. If none do, then delete that column descriptor. This check is in place so we don't have unreferenced column descriptors and columns hanging around after schema evolution for tables. This addresses bug HIVE-2246. https://issues.apache.org/jira/browse/HIVE-2246 Diffs (updated) - trunk/metastore/scripts/upgrade/derby/008-HIVE-2246.derby.sql PRE-CREATION trunk/metastore/scripts/upgrade/derby/008-REVERT-HIVE-2246.derby.sql PRE-CREATION trunk/metastore/scripts/upgrade/mysql/008-HIVE-2246.mysql.sql PRE-CREATION trunk/metastore/src/java/org/apache/hadoop/hive/metastore/HiveMetaStore.java 1153927 trunk/metastore/src/java/org/apache/hadoop/hive/metastore/ObjectStore.java 1153927 trunk/metastore/src/model/org/apache/hadoop/hive/metastore/model/MColumnDescriptor.java PRE-CREATION trunk/metastore/src/model/org/apache/hadoop/hive/metastore/model/MStorageDescriptor.java 1153927 trunk/metastore/src/model/package.jdo 1153927 trunk/ql/src/java/org/apache/hadoop/hive/ql/exec/DDLTask.java 1153927 trunk/ql/src/java/org/apache/hadoop/hive/ql/metadata/MetaDataFormatUtils.java 1153927 Diff: https://reviews.apache.org/r/1183/diff Testing --- Passes facebook's regression testing and all existing test cases. In one instance, before migration, the overhead involved with storage descriptors and columns was ~11 GB. After migration, the overhead was ~1.5 GB. Thanks, Sohan
Re: Review Request: HIVE-2246: Dedupe tables' column schemas from partitions in the metastore db
--- This is an automatically generated e-mail. To reply, visit: https://reviews.apache.org/r/1183/ --- (Updated 2011-08-08 21:19:06.999293) Review request for hive, Ning Zhang and Paul Yang. Changes --- revised description for latest changes Summary (updated) --- This patch tries to make minimal changes to the API while keeping migration short and somewhat easy to revert. The new schema can be described as follows: - CDS is a table corresponding to Column Descriptor objects. Currently, it only stores a CD_ID. - COLUMNS_V2 is a table corresponding to MFieldSchema objects, or columns. A Column Descriptor holds a list of columns. COLUMNS_V2 has a foreign key to the CD_ID to which it belongs. - SDS was modified to reference a Column Descriptor. So SDS now has a foreign key to a CD_ID which describes its columns. During migration, we create Column Descriptors for tables in a straightforward manner: their columns are now just wrapped inside a column descriptor. The SDS of partitions use their parent table's column descriptor, since currently a partition and its table share the same list of columns. When altering or adding a partition, give it it's parent table's column descriptor IF the columns they describe are the same. Otherwise, create a new column descriptor for its columns. When creating a table, create a new column descriptor every time. When altering a table, only construct a new column descriptor if the columns list has changed. Whenever you drop a storage descriptor (e.g, when dropping tables or partitions), check to see if the related column descriptor has any other references in the table. That is, check to see if any other storage descriptors point to that column descriptor. If none do, then delete that column descriptor. This check is in place so we don't have unreferenced column descriptors and columns hanging around after schema evolution for tables. This addresses bug HIVE-2246. https://issues.apache.org/jira/browse/HIVE-2246 Diffs - trunk/metastore/scripts/upgrade/derby/008-HIVE-2246.derby.sql PRE-CREATION trunk/metastore/scripts/upgrade/derby/008-REVERT-HIVE-2246.derby.sql PRE-CREATION trunk/metastore/scripts/upgrade/mysql/008-HIVE-2246.mysql.sql PRE-CREATION trunk/metastore/src/java/org/apache/hadoop/hive/metastore/HiveMetaStore.java 1153927 trunk/metastore/src/java/org/apache/hadoop/hive/metastore/ObjectStore.java 1153927 trunk/metastore/src/model/org/apache/hadoop/hive/metastore/model/MColumnDescriptor.java PRE-CREATION trunk/metastore/src/model/org/apache/hadoop/hive/metastore/model/MStorageDescriptor.java 1153927 trunk/metastore/src/model/package.jdo 1153927 trunk/ql/src/java/org/apache/hadoop/hive/ql/exec/DDLTask.java 1153927 trunk/ql/src/java/org/apache/hadoop/hive/ql/metadata/MetaDataFormatUtils.java 1153927 Diff: https://reviews.apache.org/r/1183/diff Testing --- Passes facebook's regression testing and all existing test cases. In one instance, before migration, the overhead involved with storage descriptors and columns was ~11 GB. After migration, the overhead was ~1.5 GB. Thanks, Sohan
Re: Review Request: HIVE-2246: Dedupe tables' column schemas from partitions in the metastore db
--- This is an automatically generated e-mail. To reply, visit: https://reviews.apache.org/r/1183/ --- (Updated 2011-08-08 21:29:23.722825) Review request for hive, Ning Zhang and Paul Yang. Changes --- Revert the changes to describe table T partition P, so that it always shows the table T's schema. If a table's schema has changed, we do not support querying on the old partition's schema at the moment. Summary --- This patch tries to make minimal changes to the API while keeping migration short and somewhat easy to revert. The new schema can be described as follows: - CDS is a table corresponding to Column Descriptor objects. Currently, it only stores a CD_ID. - COLUMNS_V2 is a table corresponding to MFieldSchema objects, or columns. A Column Descriptor holds a list of columns. COLUMNS_V2 has a foreign key to the CD_ID to which it belongs. - SDS was modified to reference a Column Descriptor. So SDS now has a foreign key to a CD_ID which describes its columns. During migration, we create Column Descriptors for tables in a straightforward manner: their columns are now just wrapped inside a column descriptor. The SDS of partitions use their parent table's column descriptor, since currently a partition and its table share the same list of columns. When altering or adding a partition, give it it's parent table's column descriptor IF the columns they describe are the same. Otherwise, create a new column descriptor for its columns. When creating a table, create a new column descriptor every time. When altering a table, only construct a new column descriptor if the columns list has changed. Whenever you drop a storage descriptor (e.g, when dropping tables or partitions), check to see if the related column descriptor has any other references in the table. That is, check to see if any other storage descriptors point to that column descriptor. If none do, then delete that column descriptor. This check is in place so we don't have unreferenced column descriptors and columns hanging around after schema evolution for tables. This addresses bug HIVE-2246. https://issues.apache.org/jira/browse/HIVE-2246 Diffs (updated) - trunk/metastore/scripts/upgrade/derby/008-HIVE-2246.derby.sql PRE-CREATION trunk/metastore/scripts/upgrade/derby/008-REVERT-HIVE-2246.derby.sql PRE-CREATION trunk/metastore/scripts/upgrade/mysql/008-HIVE-2246.mysql.sql PRE-CREATION trunk/metastore/src/java/org/apache/hadoop/hive/metastore/HiveMetaStore.java 1153927 trunk/metastore/src/java/org/apache/hadoop/hive/metastore/ObjectStore.java 1153927 trunk/metastore/src/model/org/apache/hadoop/hive/metastore/model/MColumnDescriptor.java PRE-CREATION trunk/metastore/src/model/org/apache/hadoop/hive/metastore/model/MStorageDescriptor.java 1153927 trunk/metastore/src/model/package.jdo 1153927 Diff: https://reviews.apache.org/r/1183/diff Testing --- Passes facebook's regression testing and all existing test cases. In one instance, before migration, the overhead involved with storage descriptors and columns was ~11 GB. After migration, the overhead was ~1.5 GB. Thanks, Sohan
Re: Review Request: HIVE-2246: Dedupe tables' column schemas from partitions in the metastore db
On 2011-07-25 06:46:04, Ning Zhang wrote: trunk/metastore/src/java/org/apache/hadoop/hive/metastore/ObjectStore.java, line 1752 https://reviews.apache.org/r/1183/diff/2/?file=26825#file26825line1752 here do you check if the 'alter table' command changes the schema (columns definition)? If it just set a table property, then you don't need to create a new ColumnDescriptor right? Also if a table's schema got changed, a new CD will be created, but the old partition will still have the old CDs. When we query the old partition, do we use the old partitons's CD or the table's CD? Also in the above case, when you run 'desc table partition old_partition', do you return the old partition's CD or the table's CD? Sohan Jain wrote: Good point; I should check whether the table columns have changed; I do this already when altering partitions. I added that in the next diff. If a table's schema changes, it does not update existing partition CDs. If we ever grab the partition object after the schema change, it will refer to its old CD, not the table's CD. However, when querying tables on the CLI, we almost always use the table's set of columns. E.g., if did: create table test (a string) partitioned by (p1 string, p2 string); alter table test add partition(p1=1, p2=1); # populate the p1=1, p2=1 partition with some data now alter table test add columns (b string) select * from test where p1 = 1 and p2 = 1, it'd use the table's latest schema; i.e., return the column 'a's values and the column 'b' as all NULL. Also, I fixed the desc table partition to use the partition's column schema, not the table's. - Sohan --- This is an automatically generated e-mail. To reply, visit: https://reviews.apache.org/r/1183/#review1176 --- On 2011-07-22 05:30:29, Sohan Jain wrote: --- This is an automatically generated e-mail. To reply, visit: https://reviews.apache.org/r/1183/ --- (Updated 2011-07-22 05:30:29) Review request for hive, Ning Zhang and Paul Yang. Summary --- This patch tries to make minimal changes to the API while keeping migration short and somewhat easy to revert. The new schema can be described as follows: - CDS is a table corresponding to Column Descriptor objects. Currently, it only stores a CD_ID. - COLUMNS_V2 is a table corresponding to MFieldSchema objects, or columns. A Column Descriptor holds a list of columns. COLUMNS_V2 has a foreign key to the CD_ID to which it belongs. - SDS was modified to reference a Column Descriptor. So SDS now has a foreign key to a CD_ID which describes its columns. During migration, we create Column Descriptors for tables in a straightforward manner: their columns are now just wrapped inside a column descriptor. The SDS of partitions use their parent table's column descriptor, since currently a partition and its table share the same list of columns. When altering or adding a partition, give it it's parent table's column descriptor IF the columns they describe are the same. Otherwise, create a new column descriptor for its columns. When adding or altering a table, create a new column descriptor every time. Whenever you drop a storage descriptor (e.g, when dropping tables or partitions), check to see if the related column descriptor has any other references in the table. That is, check to see if any other storage descriptors point to that column descriptor. If none do, then delete that column descriptor. This check is in place so we don't have unreferenced column descriptors and columns hanging around after schema evolution for tables. This addresses bug HIVE-2246. https://issues.apache.org/jira/browse/HIVE-2246 Diffs - trunk/metastore/scripts/upgrade/mysql/008-HIVE-2246.mysql.sql PRE-CREATION trunk/metastore/src/java/org/apache/hadoop/hive/metastore/ObjectStore.java 1148945 trunk/metastore/src/model/org/apache/hadoop/hive/metastore/model/MColumnDescriptor.java PRE-CREATION trunk/metastore/src/model/org/apache/hadoop/hive/metastore/model/MStorageDescriptor.java 1148945 trunk/metastore/src/model/package.jdo 1148945 Diff: https://reviews.apache.org/r/1183/diff Testing --- Passes facebook's regression testing and all existing test cases. In one instance, before migration, the overhead involved with storage descriptors and columns was ~11 GB. After migration, the overhead was ~1.5 GB. Thanks, Sohan
Re: Review Request: HIVE-2246: Dedupe tables' column schemas from partitions in the metastore db
--- This is an automatically generated e-mail. To reply, visit: https://reviews.apache.org/r/1183/ --- (Updated 2011-08-05 20:48:05.144312) Review request for hive, Ning Zhang and Paul Yang. Changes --- -On alter table, only change the column descriptor if the columns have changed. -Fix desc table partition... to use the partition's column schema, not the table's Summary --- This patch tries to make minimal changes to the API while keeping migration short and somewhat easy to revert. The new schema can be described as follows: - CDS is a table corresponding to Column Descriptor objects. Currently, it only stores a CD_ID. - COLUMNS_V2 is a table corresponding to MFieldSchema objects, or columns. A Column Descriptor holds a list of columns. COLUMNS_V2 has a foreign key to the CD_ID to which it belongs. - SDS was modified to reference a Column Descriptor. So SDS now has a foreign key to a CD_ID which describes its columns. During migration, we create Column Descriptors for tables in a straightforward manner: their columns are now just wrapped inside a column descriptor. The SDS of partitions use their parent table's column descriptor, since currently a partition and its table share the same list of columns. When altering or adding a partition, give it it's parent table's column descriptor IF the columns they describe are the same. Otherwise, create a new column descriptor for its columns. When adding or altering a table, create a new column descriptor every time. Whenever you drop a storage descriptor (e.g, when dropping tables or partitions), check to see if the related column descriptor has any other references in the table. That is, check to see if any other storage descriptors point to that column descriptor. If none do, then delete that column descriptor. This check is in place so we don't have unreferenced column descriptors and columns hanging around after schema evolution for tables. This addresses bug HIVE-2246. https://issues.apache.org/jira/browse/HIVE-2246 Diffs (updated) - trunk/metastore/src/java/org/apache/hadoop/hive/metastore/HiveMetaStore.java 1153927 trunk/metastore/src/java/org/apache/hadoop/hive/metastore/ObjectStore.java 1153927 trunk/metastore/src/model/org/apache/hadoop/hive/metastore/model/MStorageDescriptor.java 1153927 trunk/metastore/src/model/package.jdo 1153927 trunk/ql/src/java/org/apache/hadoop/hive/ql/exec/DDLTask.java 1153927 trunk/ql/src/java/org/apache/hadoop/hive/ql/metadata/MetaDataFormatUtils.java 1153927 Diff: https://reviews.apache.org/r/1183/diff Testing --- Passes facebook's regression testing and all existing test cases. In one instance, before migration, the overhead involved with storage descriptors and columns was ~11 GB. After migration, the overhead was ~1.5 GB. Thanks, Sohan
Re: Review Request: HIVE-2246: Dedupe tables' column schemas from partitions in the metastore db
--- This is an automatically generated e-mail. To reply, visit: https://reviews.apache.org/r/1183/ --- (Updated 2011-08-05 20:49:19.127572) Review request for hive, Ning Zhang and Paul Yang. Changes --- -Forgot to add a few files. NOTE: this is only a temporary diff; I need to add derby and postgres migration scripts. Summary --- This patch tries to make minimal changes to the API while keeping migration short and somewhat easy to revert. The new schema can be described as follows: - CDS is a table corresponding to Column Descriptor objects. Currently, it only stores a CD_ID. - COLUMNS_V2 is a table corresponding to MFieldSchema objects, or columns. A Column Descriptor holds a list of columns. COLUMNS_V2 has a foreign key to the CD_ID to which it belongs. - SDS was modified to reference a Column Descriptor. So SDS now has a foreign key to a CD_ID which describes its columns. During migration, we create Column Descriptors for tables in a straightforward manner: their columns are now just wrapped inside a column descriptor. The SDS of partitions use their parent table's column descriptor, since currently a partition and its table share the same list of columns. When altering or adding a partition, give it it's parent table's column descriptor IF the columns they describe are the same. Otherwise, create a new column descriptor for its columns. When adding or altering a table, create a new column descriptor every time. Whenever you drop a storage descriptor (e.g, when dropping tables or partitions), check to see if the related column descriptor has any other references in the table. That is, check to see if any other storage descriptors point to that column descriptor. If none do, then delete that column descriptor. This check is in place so we don't have unreferenced column descriptors and columns hanging around after schema evolution for tables. This addresses bug HIVE-2246. https://issues.apache.org/jira/browse/HIVE-2246 Diffs (updated) - trunk/metastore/scripts/upgrade/mysql/008-HIVE-2246.mysql.sql PRE-CREATION trunk/metastore/src/java/org/apache/hadoop/hive/metastore/HiveMetaStore.java 1153927 trunk/metastore/src/java/org/apache/hadoop/hive/metastore/ObjectStore.java 1153927 trunk/metastore/src/model/org/apache/hadoop/hive/metastore/model/MColumnDescriptor.java PRE-CREATION trunk/metastore/src/model/org/apache/hadoop/hive/metastore/model/MStorageDescriptor.java 1153927 trunk/metastore/src/model/package.jdo 1153927 trunk/ql/src/java/org/apache/hadoop/hive/ql/exec/DDLTask.java 1153927 trunk/ql/src/java/org/apache/hadoop/hive/ql/metadata/MetaDataFormatUtils.java 1153927 Diff: https://reviews.apache.org/r/1183/diff Testing --- Passes facebook's regression testing and all existing test cases. In one instance, before migration, the overhead involved with storage descriptors and columns was ~11 GB. After migration, the overhead was ~1.5 GB. Thanks, Sohan
Re: Review Request: HIVE-2246: Dedupe tables' column schemas from partitions in the metastore db
--- This is an automatically generated e-mail. To reply, visit: https://reviews.apache.org/r/1183/#review1309 --- Also, can you add migration scripts for other DB's? trunk/metastore/scripts/upgrade/mysql/008-HIVE-2246.mysql.sql https://reviews.apache.org/r/1183/#comment2982 Typo trunk/metastore/src/java/org/apache/hadoop/hive/metastore/ObjectStore.java https://reviews.apache.org/r/1183/#comment2979 The check and the delete should in the same transaction, as it's possible for a reference to a CD to be created after the check but before the delete. trunk/metastore/src/java/org/apache/hadoop/hive/metastore/ObjectStore.java https://reviews.apache.org/r/1183/#comment2981 How does this drop the storage descriptor? trunk/metastore/src/model/package.jdo https://reviews.apache.org/r/1183/#comment2968 Fix indent - Paul On 2011-08-05 20:49:19, Sohan Jain wrote: --- This is an automatically generated e-mail. To reply, visit: https://reviews.apache.org/r/1183/ --- (Updated 2011-08-05 20:49:19) Review request for hive, Ning Zhang and Paul Yang. Summary --- This patch tries to make minimal changes to the API while keeping migration short and somewhat easy to revert. The new schema can be described as follows: - CDS is a table corresponding to Column Descriptor objects. Currently, it only stores a CD_ID. - COLUMNS_V2 is a table corresponding to MFieldSchema objects, or columns. A Column Descriptor holds a list of columns. COLUMNS_V2 has a foreign key to the CD_ID to which it belongs. - SDS was modified to reference a Column Descriptor. So SDS now has a foreign key to a CD_ID which describes its columns. During migration, we create Column Descriptors for tables in a straightforward manner: their columns are now just wrapped inside a column descriptor. The SDS of partitions use their parent table's column descriptor, since currently a partition and its table share the same list of columns. When altering or adding a partition, give it it's parent table's column descriptor IF the columns they describe are the same. Otherwise, create a new column descriptor for its columns. When adding or altering a table, create a new column descriptor every time. Whenever you drop a storage descriptor (e.g, when dropping tables or partitions), check to see if the related column descriptor has any other references in the table. That is, check to see if any other storage descriptors point to that column descriptor. If none do, then delete that column descriptor. This check is in place so we don't have unreferenced column descriptors and columns hanging around after schema evolution for tables. This addresses bug HIVE-2246. https://issues.apache.org/jira/browse/HIVE-2246 Diffs - trunk/metastore/scripts/upgrade/mysql/008-HIVE-2246.mysql.sql PRE-CREATION trunk/metastore/src/java/org/apache/hadoop/hive/metastore/HiveMetaStore.java 1153927 trunk/metastore/src/java/org/apache/hadoop/hive/metastore/ObjectStore.java 1153927 trunk/metastore/src/model/org/apache/hadoop/hive/metastore/model/MColumnDescriptor.java PRE-CREATION trunk/metastore/src/model/org/apache/hadoop/hive/metastore/model/MStorageDescriptor.java 1153927 trunk/metastore/src/model/package.jdo 1153927 trunk/ql/src/java/org/apache/hadoop/hive/ql/exec/DDLTask.java 1153927 trunk/ql/src/java/org/apache/hadoop/hive/ql/metadata/MetaDataFormatUtils.java 1153927 Diff: https://reviews.apache.org/r/1183/diff Testing --- Passes facebook's regression testing and all existing test cases. In one instance, before migration, the overhead involved with storage descriptors and columns was ~11 GB. After migration, the overhead was ~1.5 GB. Thanks, Sohan
Re: Review Request: HIVE-2246: Dedupe tables' column schemas from partitions in the metastore db
--- This is an automatically generated e-mail. To reply, visit: https://reviews.apache.org/r/1183/ --- (Updated 2011-08-06 01:40:49.118616) Review request for hive, Ning Zhang and Paul Yang. Changes --- -made listStorageDescriptors.. into one transaction -renamed dropStorageDescriptorCleanly to make it's functionality clearer -indents typo Summary --- This patch tries to make minimal changes to the API while keeping migration short and somewhat easy to revert. The new schema can be described as follows: - CDS is a table corresponding to Column Descriptor objects. Currently, it only stores a CD_ID. - COLUMNS_V2 is a table corresponding to MFieldSchema objects, or columns. A Column Descriptor holds a list of columns. COLUMNS_V2 has a foreign key to the CD_ID to which it belongs. - SDS was modified to reference a Column Descriptor. So SDS now has a foreign key to a CD_ID which describes its columns. During migration, we create Column Descriptors for tables in a straightforward manner: their columns are now just wrapped inside a column descriptor. The SDS of partitions use their parent table's column descriptor, since currently a partition and its table share the same list of columns. When altering or adding a partition, give it it's parent table's column descriptor IF the columns they describe are the same. Otherwise, create a new column descriptor for its columns. When adding or altering a table, create a new column descriptor every time. Whenever you drop a storage descriptor (e.g, when dropping tables or partitions), check to see if the related column descriptor has any other references in the table. That is, check to see if any other storage descriptors point to that column descriptor. If none do, then delete that column descriptor. This check is in place so we don't have unreferenced column descriptors and columns hanging around after schema evolution for tables. This addresses bug HIVE-2246. https://issues.apache.org/jira/browse/HIVE-2246 Diffs (updated) - trunk/metastore/scripts/upgrade/mysql/008-HIVE-2246.mysql.sql PRE-CREATION trunk/metastore/src/java/org/apache/hadoop/hive/metastore/HiveMetaStore.java 1153927 trunk/metastore/src/java/org/apache/hadoop/hive/metastore/ObjectStore.java 1153927 trunk/metastore/src/model/org/apache/hadoop/hive/metastore/model/MColumnDescriptor.java PRE-CREATION trunk/metastore/src/model/org/apache/hadoop/hive/metastore/model/MStorageDescriptor.java 1153927 trunk/metastore/src/model/package.jdo 1153927 trunk/ql/src/java/org/apache/hadoop/hive/ql/exec/DDLTask.java 1153927 trunk/ql/src/java/org/apache/hadoop/hive/ql/metadata/MetaDataFormatUtils.java 1153927 Diff: https://reviews.apache.org/r/1183/diff Testing --- Passes facebook's regression testing and all existing test cases. In one instance, before migration, the overhead involved with storage descriptors and columns was ~11 GB. After migration, the overhead was ~1.5 GB. Thanks, Sohan
Re: Review Request: HIVE-2246: Dedupe tables' column schemas from partitions in the metastore db
--- This is an automatically generated e-mail. To reply, visit: https://reviews.apache.org/r/1183/#review1313 --- trunk/metastore/src/java/org/apache/hadoop/hive/metastore/ObjectStore.java https://reviews.apache.org/r/1183/#comment2984 should read 1-N actually - Sohan On 2011-08-06 01:40:49, Sohan Jain wrote: --- This is an automatically generated e-mail. To reply, visit: https://reviews.apache.org/r/1183/ --- (Updated 2011-08-06 01:40:49) Review request for hive, Ning Zhang and Paul Yang. Summary --- This patch tries to make minimal changes to the API while keeping migration short and somewhat easy to revert. The new schema can be described as follows: - CDS is a table corresponding to Column Descriptor objects. Currently, it only stores a CD_ID. - COLUMNS_V2 is a table corresponding to MFieldSchema objects, or columns. A Column Descriptor holds a list of columns. COLUMNS_V2 has a foreign key to the CD_ID to which it belongs. - SDS was modified to reference a Column Descriptor. So SDS now has a foreign key to a CD_ID which describes its columns. During migration, we create Column Descriptors for tables in a straightforward manner: their columns are now just wrapped inside a column descriptor. The SDS of partitions use their parent table's column descriptor, since currently a partition and its table share the same list of columns. When altering or adding a partition, give it it's parent table's column descriptor IF the columns they describe are the same. Otherwise, create a new column descriptor for its columns. When adding or altering a table, create a new column descriptor every time. Whenever you drop a storage descriptor (e.g, when dropping tables or partitions), check to see if the related column descriptor has any other references in the table. That is, check to see if any other storage descriptors point to that column descriptor. If none do, then delete that column descriptor. This check is in place so we don't have unreferenced column descriptors and columns hanging around after schema evolution for tables. This addresses bug HIVE-2246. https://issues.apache.org/jira/browse/HIVE-2246 Diffs - trunk/metastore/scripts/upgrade/mysql/008-HIVE-2246.mysql.sql PRE-CREATION trunk/metastore/src/java/org/apache/hadoop/hive/metastore/HiveMetaStore.java 1153927 trunk/metastore/src/java/org/apache/hadoop/hive/metastore/ObjectStore.java 1153927 trunk/metastore/src/model/org/apache/hadoop/hive/metastore/model/MColumnDescriptor.java PRE-CREATION trunk/metastore/src/model/org/apache/hadoop/hive/metastore/model/MStorageDescriptor.java 1153927 trunk/metastore/src/model/package.jdo 1153927 trunk/ql/src/java/org/apache/hadoop/hive/ql/exec/DDLTask.java 1153927 trunk/ql/src/java/org/apache/hadoop/hive/ql/metadata/MetaDataFormatUtils.java 1153927 Diff: https://reviews.apache.org/r/1183/diff Testing --- Passes facebook's regression testing and all existing test cases. In one instance, before migration, the overhead involved with storage descriptors and columns was ~11 GB. After migration, the overhead was ~1.5 GB. Thanks, Sohan
Re: Review Request: HIVE-2246: Dedupe tables' column schemas from partitions in the metastore db
--- This is an automatically generated e-mail. To reply, visit: https://reviews.apache.org/r/1183/#review1176 --- trunk/metastore/scripts/upgrade/mysql/008-HIVE-2246.mysql.sql https://reviews.apache.org/r/1183/#comment2467 is the CHARSET (latin1) the same as SDS? This will require the user's comments to be in latin1 which prevents UTF chars. trunk/metastore/scripts/upgrade/mysql/008-HIVE-2246.mysql.sql https://reviews.apache.org/r/1183/#comment2466 can you also add migration script for derby? we support derby as a default metastore RDBMS as well. trunk/metastore/src/java/org/apache/hadoop/hive/metastore/ObjectStore.java https://reviews.apache.org/r/1183/#comment2468 here do you check if the 'alter table' command changes the schema (columns definition)? If it just set a table property, then you don't need to create a new ColumnDescriptor right? Also if a table's schema got changed, a new CD will be created, but the old partition will still have the old CDs. When we query the old partition, do we use the old partitons's CD or the table's CD? Also in the above case, when you run 'desc table partition old_partition', do you return the old partition's CD or the table's CD? - Ning On 2011-07-22 05:30:29, Sohan Jain wrote: --- This is an automatically generated e-mail. To reply, visit: https://reviews.apache.org/r/1183/ --- (Updated 2011-07-22 05:30:29) Review request for hive, Ning Zhang and Paul Yang. Summary --- This patch tries to make minimal changes to the API while keeping migration short and somewhat easy to revert. The new schema can be described as follows: - CDS is a table corresponding to Column Descriptor objects. Currently, it only stores a CD_ID. - COLUMNS_V2 is a table corresponding to MFieldSchema objects, or columns. A Column Descriptor holds a list of columns. COLUMNS_V2 has a foreign key to the CD_ID to which it belongs. - SDS was modified to reference a Column Descriptor. So SDS now has a foreign key to a CD_ID which describes its columns. During migration, we create Column Descriptors for tables in a straightforward manner: their columns are now just wrapped inside a column descriptor. The SDS of partitions use their parent table's column descriptor, since currently a partition and its table share the same list of columns. When altering or adding a partition, give it it's parent table's column descriptor IF the columns they describe are the same. Otherwise, create a new column descriptor for its columns. When adding or altering a table, create a new column descriptor every time. Whenever you drop a storage descriptor (e.g, when dropping tables or partitions), check to see if the related column descriptor has any other references in the table. That is, check to see if any other storage descriptors point to that column descriptor. If none do, then delete that column descriptor. This check is in place so we don't have unreferenced column descriptors and columns hanging around after schema evolution for tables. This addresses bug HIVE-2246. https://issues.apache.org/jira/browse/HIVE-2246 Diffs - trunk/metastore/scripts/upgrade/mysql/008-HIVE-2246.mysql.sql PRE-CREATION trunk/metastore/src/java/org/apache/hadoop/hive/metastore/ObjectStore.java 1148945 trunk/metastore/src/model/org/apache/hadoop/hive/metastore/model/MColumnDescriptor.java PRE-CREATION trunk/metastore/src/model/org/apache/hadoop/hive/metastore/model/MStorageDescriptor.java 1148945 trunk/metastore/src/model/package.jdo 1148945 Diff: https://reviews.apache.org/r/1183/diff Testing --- Passes facebook's regression testing and all existing test cases. In one instance, before migration, the overhead involved with storage descriptors and columns was ~11 GB. After migration, the overhead was ~1.5 GB. Thanks, Sohan
Review Request: HIVE-2246: Dedupe tables' column schemas from partitions in the metastore db
--- This is an automatically generated e-mail. To reply, visit: https://reviews.apache.org/r/1183/ --- Review request for hive, Ning Zhang and Paul Yang. Summary --- This patch tries to make minimal changes to the API while keeping migration short and somewhat easy to revert. The new schema can be described as follows: - CDS is a table corresponding to Column Descriptor objects. Currently, it only stores a CD_ID. - COLUMNS_V2 is a table corresponding to MFieldSchema objects, or columns. A Column Descriptor holds a list of columns. COLUMNS_V2 has a foreign key to the CD_ID to which it belongs. - SDS was modified to reference a Column Descriptor. So SDS now has a foreign key to a CD_ID which describes its columns. During migration, we create Column Descriptors for tables in a straightforward manner: their columns are now just wrapped inside a column descriptor. The SDS of partitions use their parent table's column descriptor, since currently a partition and its table share the same list of columns. When altering or adding a partition, give it it's parent table's column descriptor IF the columns they describe are the same. Otherwise, create a new column descriptor for its columns. When adding or altering a table, create a new column descriptor every time. Whenever you drop a storage descriptor (e.g, when dropping tables or partitions), check to see if the related column descriptor has any other references in the table. That is, check to see if any other storage descriptors point to that column descriptor. If none do, then delete that column descriptor. This check is in place so we don't have unreferenced column descriptors and columns hanging around after schema evolution for tables. This addresses bug HIVE-2246. https://issues.apache.org/jira/browse/HIVE-2246 Diffs - trunk/metastore/src/java/org/apache/hadoop/hive/metastore/ObjectStore.java 1148945 trunk/metastore/src/model/org/apache/hadoop/hive/metastore/model/MStorageDescriptor.java 1148945 trunk/metastore/src/model/package.jdo 1148945 Diff: https://reviews.apache.org/r/1183/diff Testing --- Passes facebook's regression testing and all existing test cases. In one instance, before migration, the overhead involved with storage descriptors and columns was ~11 GB. After migration, the overhead was ~1.5 GB. Thanks, Sohan
Re: Review Request: HIVE-2246: Dedupe tables' column schemas from partitions in the metastore db
--- This is an automatically generated e-mail. To reply, visit: https://reviews.apache.org/r/1183/ --- (Updated 2011-07-22 05:30:29.026246) Review request for hive, Ning Zhang and Paul Yang. Changes --- Adding some files I missed in the last diff. Summary --- This patch tries to make minimal changes to the API while keeping migration short and somewhat easy to revert. The new schema can be described as follows: - CDS is a table corresponding to Column Descriptor objects. Currently, it only stores a CD_ID. - COLUMNS_V2 is a table corresponding to MFieldSchema objects, or columns. A Column Descriptor holds a list of columns. COLUMNS_V2 has a foreign key to the CD_ID to which it belongs. - SDS was modified to reference a Column Descriptor. So SDS now has a foreign key to a CD_ID which describes its columns. During migration, we create Column Descriptors for tables in a straightforward manner: their columns are now just wrapped inside a column descriptor. The SDS of partitions use their parent table's column descriptor, since currently a partition and its table share the same list of columns. When altering or adding a partition, give it it's parent table's column descriptor IF the columns they describe are the same. Otherwise, create a new column descriptor for its columns. When adding or altering a table, create a new column descriptor every time. Whenever you drop a storage descriptor (e.g, when dropping tables or partitions), check to see if the related column descriptor has any other references in the table. That is, check to see if any other storage descriptors point to that column descriptor. If none do, then delete that column descriptor. This check is in place so we don't have unreferenced column descriptors and columns hanging around after schema evolution for tables. This addresses bug HIVE-2246. https://issues.apache.org/jira/browse/HIVE-2246 Diffs (updated) - trunk/metastore/scripts/upgrade/mysql/008-HIVE-2246.mysql.sql PRE-CREATION trunk/metastore/src/java/org/apache/hadoop/hive/metastore/ObjectStore.java 1148945 trunk/metastore/src/model/org/apache/hadoop/hive/metastore/model/MColumnDescriptor.java PRE-CREATION trunk/metastore/src/model/org/apache/hadoop/hive/metastore/model/MStorageDescriptor.java 1148945 trunk/metastore/src/model/package.jdo 1148945 Diff: https://reviews.apache.org/r/1183/diff Testing --- Passes facebook's regression testing and all existing test cases. In one instance, before migration, the overhead involved with storage descriptors and columns was ~11 GB. After migration, the overhead was ~1.5 GB. Thanks, Sohan
Review Request: HIVE-2246: Dedupe tables' column schemas from partitions in the metastore db
--- This is an automatically generated e-mail. To reply, visit: https://reviews.apache.org/r/985/ --- Review request for hive. Summary --- We can re-organize the JDO models to reduce space usage to keep the metastore scalable for the future. Currently, partitions are the fastest growing objects in the metastore, and the metastore keeps a separate copy of the columns list for each partition. We can normalize the metastore db by decoupling Columns from Storage Descriptors and not storing duplicate lists of the columns for each partition. An idea is to create an additional level of indirection with a Column Descriptor that has a list of columns. A table has a reference to its latest Column Descriptor (note: a table may have more than one Column Descriptor in the case of schema evolution). Partitions and Indexes can reference the same Column Descriptors as their parent table. Currently, the COLUMNS table in the metastore has roughly (number of partitions + number of tables) * (average number of columns pertable) rows. We can reduce this to (number of tables) * (average number of columns per table) rows, while incurring a small cost proportional to the number of tables to store the Column Descriptors. This addresses bug HIVE-2246. https://issues.apache.org/jira/browse/HIVE-2246 Diffs - trunk/metastore/if/hive_metastore.thrift 1140399 trunk/metastore/src/model/org/apache/hadoop/hive/metastore/model/MColumnDescriptor.java PRE-CREATION trunk/metastore/src/model/org/apache/hadoop/hive/metastore/model/MDatabase.java 1140399 trunk/metastore/src/model/org/apache/hadoop/hive/metastore/model/MFieldSchema.java 1140399 trunk/metastore/src/model/org/apache/hadoop/hive/metastore/model/MIndex.java 1140399 trunk/metastore/src/model/org/apache/hadoop/hive/metastore/model/MPartition.java 1140399 trunk/metastore/src/model/org/apache/hadoop/hive/metastore/model/MStorageDescriptor.java 1140399 trunk/metastore/src/model/org/apache/hadoop/hive/metastore/model/MTable.java 1140399 trunk/metastore/src/model/package.jdo 1140399 trunk/ql/src/java/org/apache/hadoop/hive/ql/exec/DDLTask.java 1140399 trunk/ql/src/java/org/apache/hadoop/hive/ql/index/TableBasedIndexHandler.java 1140399 trunk/ql/src/java/org/apache/hadoop/hive/ql/index/bitmap/BitmapIndexHandler.java 1140399 trunk/ql/src/java/org/apache/hadoop/hive/ql/index/compact/CompactIndexHandler.java 1140399 trunk/ql/src/java/org/apache/hadoop/hive/ql/metadata/Hive.java 1140399 trunk/ql/src/java/org/apache/hadoop/hive/ql/metadata/MetaDataFormatUtils.java 1140399 trunk/ql/src/java/org/apache/hadoop/hive/ql/metadata/Partition.java 1140399 trunk/ql/src/java/org/apache/hadoop/hive/ql/metadata/Table.java 1140399 trunk/ql/src/java/org/apache/hadoop/hive/ql/parse/ImportSemanticAnalyzer.java 1140399 Diff: https://reviews.apache.org/r/985/diff Testing --- Haven't run any unit tests yet, just qualitative testing so far. Thanks, Sohan