Uh oh!
There was an error while loading. Please reload this page.
TEZ-4654: Migrate from commons-lang2.x to commons-lang3.x - #441
Conversation
Aggarwal-Raghav
commented
Nov 12, 2025
This comment was marked as outdated.
This comment was marked as outdated.
25ee1ed to
cafecd5CompareAggarwal-Raghav
commented
Nov 12, 2025
Looks like UT failures are because of exclusion of common-lang 2.x in From Hive perspective, commons-lang 2.x is part of hive-exec jar so, it shouldn't cause any runtime issues as well. https://github.com/apache/hive/blob/c054a47c7c22875c93ca0d29b7ac730cf7deba04/ql/pom.xml#L1060 CC @abstractdog |
This comment was marked as outdated.
This comment was marked as outdated.
@Aggarwal-Raghav : I can see only 1 problem in tez-javadoc-tools while compiling the java 17+ version |
@abstractdog , it is passing in local. Attaching the build logs for this module. Do i need to force push again? |
Aggarwal-Raghav
commented
Dec 2, 2025
Please suggest, what needs to be done. |
sorry for the late reply, it has passed for me too locally, can you rebase and push to see it now? and then we'll decide |
- Upgrade to commons-lang-3.19.0 - org.apache.commons.lang.ArrayUtils => org.apache.commons.lang3.ArrayUtils - org.apache.commons.lang.RandomStringUtils => org.apache.commons.lang3.RandomStringUtils - org.apache.commons.lang.StringEscapeUtils => org.apache.commons.lang3.StringEscapeUtils - org.apache.commons.lang.StringUtils => org.apache.commons.lang3.StringUtils - org.apache.commons.lang.SystemUtils => org.apache.commons.lang3.SystemUtils - org.apache.commons.lang.exception.ExceptionUtils => org.apache.commons.lang3.exception.ExceptionUtils - org.apache.commons.lang.mutable.MutableInt => org.apache.commons.lang3.mutable.MutableInt - org.apache.commons.lang.NotImplementedException => org.apache.commons.lang3.NotImplementedException
cafecd5 to
c884d63CompareAggarwal-Raghav
commented
Jan 22, 2026
No Worries. Rebased |
tez-yetus
commented
Jan 22, 2026
💔 -1 overall
This message was automatically generated. |
abstractdog
commented
Jan 22, 2026
looks good to me, spotbugs should be fixed though in TEZ-4669 |
abstractdog
commented
Jan 22, 2026
I'll double-check the javadoc tools build, because it's weird...it compiles for me locally too |
Aggarwal-Raghav
commented
Jan 22, 2026
I'm speculating and might be wrong but is it possible that the |
Aggarwal-Raghav
commented
Jan 22, 2026
As its not able to find the previously build modules |
but right before these modules' build, there is e.g. tez-dag build as shown in which ends with: I cannot reproduce this locally now, and now it's not only tez-javadoc-tools but also tez-tests more interestingly, tez-runtime-internals builds successfully, even though it's complaining about tez-api, which it depends on: |
abstractdog
commented
Jan 22, 2026
I just hacked yetus to add -X to maven command and explode, will see :) |
tez-yetus
commented
Jan 22, 2026
💔 -1 overall
This message was automatically generated. |
maheshrajus
commented
Jan 23, 2026
@Aggarwal-Raghav spotbugs report showing -1 for all modules. It is failed due to below error. Can you check it ? |
abstractdog
commented
Jan 23, 2026
already tracked on TEZ-4669 |
debug level maven logs revealed new exotic things: so invalid POM is some missing version, but still don't get how could this come in precommit and not locally |
reproduced the maven compile problem on an ubuntu server by hitting the same commands as yetus: again, the same didn't happen on my mac |
Aggarwal-Raghav
commented
Jan 23, 2026
I'll also look into this over the weekend. |
Aggarwal-Raghav
commented
Jan 23, 2026
Can we remove this |
abstractdog
commented
Jan 23, 2026
I'm afraid we cannot, this is deep inside yetus logic, how it identifies the affected modules and their dependencies, and build them one by one |
abstractdog
commented
Jan 23, 2026
ignore my last comment about effective pom, it was wrong, deleted it |
hm, strange, now I can repro the same locally too :D |
I'm also able to repro with the steps you have shared on mac but @abstractdog , isn't this way of doing things wrong? if clean the ~/.m2/repository cache and build the modules like |
Aggarwal-Raghav
commented
Jan 23, 2026
That's what I was suspecting yesterday as well. Maybe I couldn't communicate it better. |
I agree that building the project module by module is not necessarily the most optimal way, because we can figure out what went wrong even if the whole project is built in one go, I'm totally fine with a patch where the maven build phase is done in one go, and as a positive side effect, this issue would also go away |
made a decision here, this PR should be merged, I'm not about to block valuable contributions because we cannot figure out and solve infra problems easily, so in this case:
this patch looks safe and clean |
Uh oh!
There was an error while loading. Please reload this page.

Uh oh!
There was an error while loading. Please reload this page.