This is most likely due to cleanup not happening properly from the previous
runs.
On 6/28/18, 10:48 AM, "Vineet Garg" wrote:
I have also bumped into flaky druid failures although with different
symptoms (Wait time exhausted and we have [1] out of [1] segments not loaded
yet)
All
I have also bumped into flaky druid failures although with different symptoms
(Wait time exhausted and we have [1] out of [1] segments not loaded yet)
All of the druid failures in this case are during create table with following
error stack: (Not sure why this is intermittent)
org.apache.hadoop
Ptests have become really unstable.
The druid tests are failing randomly,
https://builds.apache.org/job/PreCommit-HIVE-Build/12203/testReport
Should we disable them?
Deepak
On 6/27/18, 10:13 AM, "Deepak Jaiswal" wrote:
Hi All,
It seems we are going back to instability in Hive QA
Hi All,
It seems we are going back to instability in Hive QA runs. In the past few days
I saw many runs where the failures were completely independent. When those
tests are run locally, they don’t fail which makes them harder to catch.
On one side I think having green run to commit makes sense,
