Nicholas Jiang
Nicholas Jiang
@mridulm, PTAL.
@mridulm, @pan3793, @cfmcgrady, @cxzl25, I have address above comments. PTAL.
@mridulm, I just tested with Spark SQL via above command. Did you ever try to test Spark SQL to verify? Meanwhile, why does the above application fail with `Caused by:...
@mridulm, my deployment experience is that the celeborn client jar is not placed in `jars` directory but placed in HDFS which path is set to `spark.jars` for using `SparkShuffleManager`. The...
@mridulm, I uses the internal Spark 3.1 version to verify this. Anyway, this pull request could help user to confirm whether to use Celeborn in `SparkShuffleManager`. @mridulm, @pan3793, @cfmcgrady, @cxzl25,...
Merged to main(v0.5.0).
Ping @FMX, @szetszwo, @pan3793, @cxzl25.
@szetszwo, thanks for your review. cc @FMX.
@pan3793, the rolling upgrade strategy should be: 1. Try to upgrade a non-leader master node first. 2. Follow the strategy of upgrading a single worker node - multiple worker nodes...
@FMX, I didn't test the rolling upgrade process in cluster. I would like to try rolling upgrade for validation of above question.