[Doc] Update 22.08 documentation

[viadea]

Update the 22.08 documentation.

1. Add release notes for 22.08
2. Change the diagram from Spark 3.0 to Spark 3.x

Note: In this PR, I did not add the download links for 22.08 jars or modify the Tool jar links to avoid link check failure issues.
I will do it right before the final merge PR in another PR.

[sameerz]

Other changes:

• Update the FAQ to note we support Spark 3.2.2
• There is a table formatting issue on gh-pages in https://nvidia.github.io/spark-rapids/docs/compatibility.html#csv-ansi-day-time-interval, possibly due to a lack of a blank line above the text "Currently, the RAPIDS Accelerator only supports the ANSI style." in https://raw.githubusercontent.com/NVIDIA/spark-rapids/branch-22.08/docs/compatibility.md .
• In https://github.com/NVIDIA/spark-rapids/blob/branch-22.08/docs/spark-profiling-tool.md , change Profiling tool to Profiling Tool to be consistent with other page titles at https://nvidia.github.io/spark-rapids/

[sameerz]

from_json will be in 22.10 ( PR #6211 )

[viadea]

Removed.

[viadea]

Fixed all above.

[sameerz]

Can we update https://github.com/NVIDIA/spark-rapids/blob/branch-22.08/docs/configs.md (which is a generated file) to remove --conf spark.rapids.sql.incompatibleOps.enabled=true since it is now true by default?

[nvliyuan]

if a customer setup a databricks cluster, then log in to one cluster node and run some codes through pyspark/spark-shell scripts with rapids-plugin, it will complain that the spark shim loader can not be found, because it is up to spark.databricks.clusterUsageTags.sparkVersion to decide which databricks shim should be load, so customers need to set an additional config --conf spark.rapids.shims-provider-override=com.nvidia.spark.rapids.shims.spark3XXdb.SparkShimServiceProvider; this is set automatically on databricks notebook but not databricks scripts, maybe we should document it in databricks getting started guide

[viadea]

Good information. But I feel Databricks users may not really want to SSH to the nodes and run spark-shell.

[viadea]

Can we update https://github.com/NVIDIA/spark-rapids/blob/branch-22.08/docs/configs.md (which is a generated file) to remove --conf spark.rapids.sql.incompatibleOps.enabled=true since it is now true by default?

@sameerz I modified below 1 files to use another parameter spark.rapids.sql.concurrentGpuTasks to replace spark.rapids.sql.incompatibleOps.enabled:
./sql-plugin/src/main/scala/com/nvidia/spark/rapids/RapidsConf.scala

Also fixed one typo : spark -> spark-shell

For ./spark2-sql-plugin/src/main/scala/com/nvidia/spark/rapids/RapidsConf.scala, Tom will use a different PR to fix it.

[sameerz]

Looks good. I am going to remove the [skip ci] since there are changes to RapidsConf.scala.

[sameerz]

build