haojinIntel

Using llvm-9 and clang-9 will encounter hanging issue when running q14a.sql on ubuntu cluster.

3

For OAP-1.4 testing, we encounter hanging issue when running q14a.sql on our ubuntu cluster.  When we set "spark.oap.sql.columnar.hashagg" equal to false, the query can pass.

Q53.sql failed on 8th rounds when running TPC-DS for 20 rounds with 1.5TB data scale.

1

We used the code of branch-1.4 and triggered stability test for gazelle and run TPC-DS with 1.5TB data scale for 20 rounds. Q53.sql failed on 8th round thanks to the...

Run Q14a.sql , q14b.sql of TPC-DS SF10TB will fail thanks to lacking of offheap memory.

1

We use the cluster with 1 master and 3 workers. Each worker contains 128 vcores and 512GB DRAM. Vanilla spark can successfully run TPC-DS F10TB while gazelle will fail during...

Run Q14a.sql , q14b.sql of TPC-DS will meet hanging issue with the setting "spark.sql.inMemoryColumnarStorage.batchSize=32768"

1

We use the settings about batch size like below confs: ``` spark.sql.inMemoryColumnarStorage.batchSize 32768 spark.sql.execution.arrow.maxRecordsPerBatch 32768 spark.sql.parquet.columnarReaderBatchSize 32768 ``` When running q14a.sql or q14b.sql, the test will be hang on native...

OAP-1.5 has performance regression comparing with OAP-1.4

1

The queries of TPC-DS with SMJ has obvious performance regression.

sbt package failed with unresolved dependency

5

I've installed sbt-0.13.15 in my environment. I trigger "sbt packge" and meet the following exceptions:  Is there anyone meet the similar issue? And how can I fix this problem.

SQL Data Source Cache Plasma Backend can not compatible with spark-sql_2.12-3.1.1-amzn-0.jar on AWS EMR-6.3.0

1

For OAP-1.2, we target to support all features on cloud. We choose emr-6.3.0 and the versions of software are showed below:  We try to run TPC-DS power test with...

Cannot easily start plasma service on AWS EMR cluster.

8

For OAP-1.2, we target to support all features on cloud. We choose emr-6.3.0 and the versions of software are showed below:  For plasma-sql-ds-cache it's not convenient to log on...

Plasma-SQL-DS-Cache of OAP-1.1 has performance regression comparing with OAP-1.1.1

2

We use the same configuration to run TPC-DS with 3TB data scale. The cluster contains 3 workers and each contains 1TB PMEM. The performance has 12.9% regression. Component | Configuration...