You are viewing a plain text version of this content. The canonical link for it is here.
Posted to issues@spark.apache.org by "Jacek Laskowski (JIRA)" <ji...@apache.org> on 2017/05/10 08:06:04 UTC

[jira] [Created] (SPARK-20691) Difference between Storage Memory as seen internally and in web UI

Jacek Laskowski created SPARK-20691:
---------------------------------------

             Summary: Difference between Storage Memory as seen internally and in web UI
                 Key: SPARK-20691
                 URL: https://issues.apache.org/jira/browse/SPARK-20691
             Project: Spark
          Issue Type: Bug
          Components: Web UI
    Affects Versions: 2.3.0
            Reporter: Jacek Laskowski


I set Major priority as it's visible to a user.

There's a difference in what the size of Storage Memory is managed internally and displayed to a user in web UI.

I found it while answering [How does web UI calculate Storage Memory (in Executors tab)?|http://stackoverflow.com/q/43801062/1305344] on StackOverflow.

In short (quoting the main parts), when you start a Spark app (say spark-shell) you see 912.3 MB RAM for Storage Memory:

{code}
$ ./bin/spark-shell --conf spark.driver.memory=2g
...
17/05/07 15:20:50 INFO BlockManagerMasterEndpoint: Registering block manager 192.168.1.8:57177 with 912.3 MB RAM, BlockManagerId(driver, 192.168.1.8, 57177, None)
{code}

but in the web UI you'll see 956.6 MB due to the way the custom JavaScript function {{formatBytes}} in [utils.js|https://github.com/apache/spark/blob/master/core/src/main/resources/org/apache/spark/ui/static/utils.js#L40-L48] calculates the value. That translates to the following Scala code:

{code}
def formatBytes(bytes: Double) = {
  val k = 1000
  val i = math.floor(math.log(bytes) / math.log(k))
  val maxMemoryWebUI = bytes / math.pow(k, i)
  f"$maxMemoryWebUI%1.1f"
}
scala> println(formatBytes(maxMemory))
956.6
{code}



--
This message was sent by Atlassian JIRA
(v6.3.15#6346)

---------------------------------------------------------------------
To unsubscribe, e-mail: issues-unsubscribe@spark.apache.org
For additional commands, e-mail: issues-help@spark.apache.org