rdsr opened a new issue #540: A Table should optionally allow for automatic
schema evolution
URL: https://github.com/apache/incubator-iceberg/issues/540
Related to #244 . We should allow for automatic schema evolution when
writing to an Iceberg table. This feature should be controlled by
rdblue commented on issue #529: Add hadoop table catalog (WIP)
URL: https://github.com/apache/incubator-iceberg/pull/529#issuecomment-541442319
@chenjunjiedada, can you update the PR description with a summary of what
you're proposing here?
If I understand correctly, the idea is to
rdsr commented on issue #540: Iceberg tables should optionally allow for
automatic schema evolution
URL:
https://github.com/apache/incubator-iceberg/issues/540#issuecomment-541442220
> It's easy to add columns with the wrong type when you're doing that.
Wouldn't `UpdateSchema`
rdblue commented on a change in pull request #526: Add Baseline to
iceberg-parquet
URL: https://github.com/apache/incubator-iceberg/pull/526#discussion_r334289512
##
File path:
parquet/src/main/java/org/apache/iceberg/parquet/ParquetSchemaUtil.java
##
@@ -29,6 +29,11 @@
rdblue commented on a change in pull request #526: Add Baseline to
iceberg-parquet
URL: https://github.com/apache/incubator-iceberg/pull/526#discussion_r334289534
##
File path:
parquet/src/main/java/org/apache/iceberg/parquet/ParquetTypeVisitor.java
##
@@ -29,9 +29,8 @@
rdblue commented on a change in pull request #526: Add Baseline to
iceberg-parquet
URL: https://github.com/apache/incubator-iceberg/pull/526#discussion_r334289657
##
File path:
parquet/src/main/java/org/apache/iceberg/parquet/ParquetValueReaders.java
##
@@ -584,9 +585,9
rdblue closed pull request #309: Add mechanism to expire table metadata
URL: https://github.com/apache/incubator-iceberg/pull/309
This is an automated message from the Apache Git Service.
To respond to the message, please
rdblue commented on issue #309: Add mechanism to expire table metadata
URL: https://github.com/apache/incubator-iceberg/pull/309#issuecomment-541449375
I'm closing this for now, since the approach is probably not what we will go
with in the end, but I'd like to get a solution for this
rdsr commented on issue #540: Iceberg tables should optionally allow for
automatic schema evolution
URL:
https://github.com/apache/incubator-iceberg/issues/540#issuecomment-541441863
Yes, if the columns are valid as per the schema evolution rules
We've seen use-cases of it where users
rdblue commented on a change in pull request #525: Apply Baseline to iceberg-pig
URL: https://github.com/apache/incubator-iceberg/pull/525#discussion_r334289942
##
File path: pig/src/main/java/org/apache/iceberg/pig/IcebergStorage.java
##
@@ -193,80 +166,81 @@ public void
rdblue commented on a change in pull request #525: Apply Baseline to iceberg-pig
URL: https://github.com/apache/incubator-iceberg/pull/525#discussion_r334290084
##
File path: pig/src/main/java/org/apache/iceberg/pig/PigParquetReader.java
##
@@ -366,16 +383,16 @@ protected
rdsr opened a new issue #541: Iceberg Pig reader should support both catalogs
URL: https://github.com/apache/incubator-iceberg/issues/541
In `IcebergStorage` we are currently only using the `Tables` api which is
the filesystem or Hadoop catalog. We should support both. Or better yet, have
rdblue commented on issue #181: Add mechanism to expire old metadata versions
URL:
https://github.com/apache/incubator-iceberg/issues/181#issuecomment-541449685
From discussion on #309, I think the right way to expire old metadata is to
keep a list of the N previous metadata files in each
rdblue commented on issue #540: Iceberg tables should optionally allow for
automatic schema evolution
URL:
https://github.com/apache/incubator-iceberg/issues/540#issuecomment-541440371
What do you mean by automatic schema evolution? When you write to a table,
any new columns are
rdblue commented on issue #540: Iceberg tables should optionally allow for
automatic schema evolution
URL:
https://github.com/apache/incubator-iceberg/issues/540#issuecomment-541442093
Yeah, I think that could be a good idea. We'd want to document some of the
risks, though. It's easy to
rdblue commented on a change in pull request #526: Add Baseline to
iceberg-parquet
URL: https://github.com/apache/incubator-iceberg/pull/526#discussion_r334289245
##
File path: parquet/src/main/java/org/apache/iceberg/parquet/Parquet.java
##
@@ -98,18 +98,18 @@ public
rdblue commented on a change in pull request #526: Add Baseline to
iceberg-parquet
URL: https://github.com/apache/incubator-iceberg/pull/526#discussion_r334289220
##
File path:
parquet/src/main/java/org/apache/iceberg/parquet/ParquetDictionaryRowGroupFilter.java
##
@@
rdsr commented on a change in pull request #525: Apply Baseline to iceberg-pig
URL: https://github.com/apache/incubator-iceberg/pull/525#discussion_r334289841
##
File path: pig/src/main/java/org/apache/iceberg/pig/IcebergStorage.java
##
@@ -290,8 +265,8 @@ public
rdsr commented on a change in pull request #525: Apply Baseline to iceberg-pig
URL: https://github.com/apache/incubator-iceberg/pull/525#discussion_r334289743
##
File path: pig/src/main/java/org/apache/iceberg/pig/SchemaUtil.java
##
@@ -73,6 +60,16 @@ private static
rdblue commented on a change in pull request #526: Add Baseline to
iceberg-parquet
URL: https://github.com/apache/incubator-iceberg/pull/526#discussion_r334289375
##
File path:
parquet/src/main/java/org/apache/iceberg/parquet/ParquetDictionaryRowGroupFilter.java
##
@@
rdsr commented on a change in pull request #525: Apply Baseline to iceberg-pig
URL: https://github.com/apache/incubator-iceberg/pull/525#discussion_r334289804
##
File path: pig/src/main/java/org/apache/iceberg/pig/IcebergPigInputFormat.java
##
@@ -29,7 +29,7 @@
import
rdsr commented on a change in pull request #525: Apply Baseline to iceberg-pig
URL: https://github.com/apache/incubator-iceberg/pull/525#discussion_r334279185
##
File path: pig/src/main/java/org/apache/iceberg/pig/PigParquetReader.java
##
@@ -366,16 +383,16 @@ protected
rdblue commented on issue #540: Iceberg tables should optionally allow for
automatic schema evolution
URL:
https://github.com/apache/incubator-iceberg/issues/540#issuecomment-541446146
Yes, it will. But we would want to fail earlier than that. The case I'm
thinking about is when you're
rdblue commented on issue #539: Replace StringBuffer by StringBuilder
URL: https://github.com/apache/incubator-iceberg/pull/539#issuecomment-541441529
Is the behavior of `setLength` the same for both classes? This section is
tricky because some of the lengths are measured in bytes, some in
rdblue commented on a change in pull request #537: Docs: Fix typos
URL: https://github.com/apache/incubator-iceberg/pull/537#discussion_r334288518
##
File path: site/docs/api-quickstart.md
##
@@ -113,7 +113,7 @@ import org.apache.avro.Schema.Parser
val avroSchema = new
rdblue commented on a change in pull request #526: Add Baseline to
iceberg-parquet
URL: https://github.com/apache/incubator-iceberg/pull/526#discussion_r334289415
##
File path: parquet/src/main/java/org/apache/iceberg/parquet/ParquetFilters.java
##
@@ -168,9 +169,9 @@
rdsr opened a new issue #542: Support other data formats in Iceberg Pig reader
URL: https://github.com/apache/incubator-iceberg/issues/542
Today `IcebergRecordReader` only supports Parquet
This is an automated message from
rdblue commented on a change in pull request #525: Apply Baseline to iceberg-pig
URL: https://github.com/apache/incubator-iceberg/pull/525#discussion_r334290142
##
File path: build.gradle
##
@@ -451,6 +451,8 @@ project(':iceberg-pig') {
compile
28 matches
Mail list logo