Github user squito commented on a diff in the pull request:
https://github.com/apache/spark/pull/22504#discussion_r226400385
--- Diff: docs/configuration.md ---
@@ -266,6 +266,37 @@ of the most common options to set are:
Only has effect in Spark standalone mode or Mesos cluster deploy mode.
</td>
</tr>
+<tr>
+ <td><code>spark.driver.log.dfsDir</code></td>
+ <td>(none)</td>
+ <td>
+ Base directory in which Spark driver logs are synced, if
spark.driver.log.syncToDfs.enabled is true.
+ Within this base directory, Spark creates a sub-directory for each
application, and logs the driver logs
+ specific to the application in this directory. Users may want to set
this to a unified location like an
+ HDFS directory so driver log files can be persisted for later usage.
This directory should allow any spark
+ user to read/write files and the spark history server user to delete
files. Additionally, older logs from
+ this directory are cleaned by Spark History Server if
spark.history.fs.driverlog.cleaner.enabled is true.
+ They are cleaned if they are older than max age configured at
spark.history.fs.driverlog.cleaner.maxAge.
+ </td>
+</tr>
+<tr>
+ <td><code>spark.driver.log.syncToDfs.enabled</code></td>
+ <td>false</td>
+ <td>
+ If true, spark application running in client mode will sync driver
logs to a persistent storage, configured
--- End diff --
it seems like there is a mismatch between what is described here and what
is implemented. Do you intend to support a configuration where you are only
logging to local disk, and not syncing to dfs? It doesn't seem like
`DriverLogger.apply` allows that. And if you do intend to support that, and
want it controlled by this configuration, then I'd remove the "syncToDfs" from
this name.
---
---------------------------------------------------------------------
To unsubscribe, e-mail: reviews-unsubscr...@spark.apache.org
For additional commands, e-mail: reviews-h...@spark.apache.org
