[SPARK-17558] Bump Hadoop 2.7 version from 2.7.2 to 2.7.3

[rxin]

What changes were proposed in this pull request?

This patch bumps the Hadoop version in hadoop-2.7 profile from 2.7.2 to 2.7.3, which was recently released and contained a number of bug fixes.

How was this patch tested?

The change should be covered by existing tests.

[rxin]

cc @JoshRosen @srowen

[SparkQA]

Test build #65470 has finished for PR 15115 at commit 1fc9047.

• This patch passes all tests.
• This patch merges cleanly.
• This patch adds no public classes.

[srowen]

+1, can't hurt

[rxin]

Merging in master/2.0.

[steveloughran]

This seems a duplicate of the #14827 patch I filed 3 weeks earlier.

Is there some aspect of the PR submission process that I'm missing out on? I would like to get my patches in, and having duplicate works wastes everyone's time.

What am I doing wrong?

[srowen]

The only difference was that you're also making 2.7 the default, which isn't bad or anything. Otherwise i think it was just an oversight. There's way too much traffic to keep track of unfortunately, which is why I encourage steps to reduce the noise on lists/JIRA/PRs.

You have a couple WIP PRs you've bene updating for a while so I also sort of look past yours unfortunately in the email flood as I think they're just WIP updates.

[steveloughran]

thanks for clarifying; sometimes I feel that my patches get under reviewed —which holds for Hadoop too, where some have been outstanding for so long they're approach school age

[srowen]

#14731 is waiting on your response if I read it correctly; #12004 is a WIP (not sure if that's going to get merged anyway); #9571 seems to have some comments from Marcelo; #14646 I'm not sure is necessary -- was it to be committed? #13579 has a question for you from Marcelo. I'm not sure you're being ignored?