Marcel Kornacker has posted comments on this change. Change subject: IMPALA-5338: Fix Kudu timestamp column default values ......................................................................
Patch Set 1: (4 comments) http://gerrit.cloudera.org:8080/#/c/6936/1//COMMIT_MSG Commit Message: Line 19: will be a BIGINT if the column type is TIMESTAMP. It is only what if you stored the default as a timestamp literal instead (and then convert it to what kudu needs when that's required)? it feels like the current approach requires a bunch of special-casing. http://gerrit.cloudera.org:8080/#/c/6936/1/fe/src/main/java/org/apache/impala/analysis/ColumnDef.java File fe/src/main/java/org/apache/impala/analysis/ColumnDef.java: Line 87: // post-analysis. For TIMESTAMPs those are Kudu UNIXTIME_MICROS, i.e. int64s. i don't like that this changes its meaning as a result of analysis, because we're planning on making analysis idempotent (at some point in the future). let's make it a separate variable, and then you can get rid of isanalyzed_ http://gerrit.cloudera.org:8080/#/c/6936/1/fe/src/main/java/org/apache/impala/util/KuduUtil.java File fe/src/main/java/org/apache/impala/util/KuduUtil.java: Line 167: public static Object getKuduDefaultValue(TExpr defaultValue, supply function comment http://gerrit.cloudera.org:8080/#/c/6936/1/fe/src/test/java/org/apache/impala/analysis/AnalyzeDDLTest.java File fe/src/test/java/org/apache/impala/analysis/AnalyzeDDLTest.java: Line 2403: "ts timestamp not null default cast('2009-01-01 00:00:00' as timestamp)) " + should we also allow a string constant here (that we then try to convert to a timestamp, and if that fails, we complain)? -- To view, visit http://gerrit.cloudera.org:8080/6936 To unsubscribe, visit http://gerrit.cloudera.org:8080/settings Gerrit-MessageType: comment Gerrit-Change-Id: I655910fb4805bb204a999627fa9f68e43ea8aaf2 Gerrit-PatchSet: 1 Gerrit-Project: Impala-ASF Gerrit-Branch: master Gerrit-Owner: Matthew Jacobs <m...@cloudera.com> Gerrit-Reviewer: Alex Behm <alex.b...@cloudera.com> Gerrit-Reviewer: Dimitris Tsirogiannis <dtsirogian...@cloudera.com> Gerrit-Reviewer: Marcel Kornacker <mar...@cloudera.com> Gerrit-Reviewer: Matthew Jacobs <m...@cloudera.com> Gerrit-Reviewer: Thomas Tauber-Marshall <tmarsh...@cloudera.com> Gerrit-HasComments: Yes