Sherif Ebady

We've been dealing with this "issue" too. It's not at all clear that an image named `gcr.io/spark-operator/spark:v3.1.1` would be using Spark 3 with Hadoop 2.7.x libraries. Ideally the image gcr.io/spark-operator/spark:v3.1.1...

The self healing is fully aware of when the browsers / tests are running and therefore is the ideal place to cleanup these tmp files that are related to browsers.

You can monitor /tmp but I think a simpler cleanup process is possible. Wait till all tests have finished/stopped and browsers terminated during self healing then remove these directories: /tmp/anonymous*webdriver-profile...

I wanted to look at this and see if it is possible to support this. If not then a better exception might be called for here.

This issue was opened from https://github.com/paypal/squbs/issues/519. Should also consider optimization to avoid scheduleOnce if timeLeftForNextElemToTimeout is near the timer precision.

Another failure here: 08:24:10 [info] - should handle request with empty web-context 08:24:11 08:24:11.005 [pool-1-thread-1] INFO org.squbs.unicomplex.UnicomplexBoot$ - Started cube org.squbs.unicomplex.test.MultiListener 0.0.1 08:24:11 08:24:11.031 [pool-1-thread-1] INFO org.squbs.unicomplex.UnicomplexBoot$ - Web Service...

Another travis failure https://api.travis-ci.org/v3/job/485606748/log.txt [info] - should recover from downstream failure 08:00:23.408 [pool-1-thread-1-ScalaTest-running-BroadcastBooleanBufferNoAutoCommitSpec] DEBUG net.openhft.chronicle.bytes.MappedFile - Allocation of 0 chunk in /tmp/persistent_queue2046121375657204605/directory-listing.cq4t took 0.1 ms. 08:00:23.410 [pool-1-thread-1-ScalaTest-running-BroadcastBooleanBufferNoAutoCommitSpec] DEBUG net.openhft.chronicle.bytes.MappedFile -...

I have a use case for using a custom SMT with MM2.