spark-user mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Mark Hamstra <m...@clearstorydata.com>
Subject Re: "Loading" status
Date Mon, 02 Feb 2015 17:33:39 GMT
Yes, if the Master is unable to register the Executor and transition it to
RUNNING, then the Executor will stay in LOADING state, so this can be
caused by problems in the Master or the Master-Executor communication.


On Mon, Feb 2, 2015 at 9:24 AM, Tushar Sharma <tusharshar@gmail.com> wrote:

> Yes curious indeed. Usually if such status persists it leads to executor
> failure after a couple of tries. We also noticed similar behavior but the
> logs clearly mentioned the lack of some akka resource as the reason. I am
> sure the logs also holds the key for this particular problem.
>
> On Mon, Feb 2, 2015 at 10:49 PM, Mark Hamstra <mark@clearstorydata.com>
> wrote:
>
>> Curious.  I guess the first question is whether we've got some sort of
>> Listener/UI error so that the UI is not accurately reflecting the
>> Executor's actual state, or whether the "LOADING" Executor really is
>> spending a considerable length of time in this "I'm in the process of being
>> created, but not yet doing anything useful" state.
>>
>> If you can figure out a little more of what is going on or how to
>> reproduce this state, please do file a JIRA.
>>
>>
>> On Mon, Feb 2, 2015 at 8:28 AM, Ami Khandeshi <ami.khandeshi@gmail.com>
>> wrote:
>>
>>> Yes
>>>
>>>
>>> On Monday, February 2, 2015, Mark Hamstra <mark@clearstorydata.com>
>>> wrote:
>>>
>>>> LOADING is just the state in which new Executors are created but before
>>>> they have everything they need and are fully registered to transition to
>>>> state RUNNING and begin doing actual work:
>>>>
>>>> https://github.com/apache/spark/blob/master/core/src/main/scala/org/apache/spark/deploy/worker/Worker.scala#L351
>>>>
>>>> https://github.com/apache/spark/blob/master/core/src/main/scala/org/apache/spark/deploy/client/AppClient.scala#L133
>>>>
>>>> LOADING should be a fairly brief, transitory state.  Are you seeing
>>>> Executors remaining in LOADING for a significant length of time?
>>>>
>>>>
>>>> On Mon, Feb 2, 2015 at 7:56 AM, akhandeshi <ami.khandeshi@gmail.com>
>>>> wrote:
>>>>
>>>>> I am not sure what Loading status means, followed by Running.  In the
>>>>> application UI, I see:
>>>>> Executor Summary
>>>>>
>>>>> ExecutorID      Worker  Cores   Memory  State   Logs
>>>>> 1       worker-20150202144112-hadoop-w-1.c.fi-mdd-poc.internal-38874
>>>>>   16      83971
>>>>> LOADING stdout stderr
>>>>> 0       worker-20150202144112-hadoop-w-2.c.fi-mdd-poc.internal-58685
>>>>>   16      83971
>>>>> RUNNING stdout stderr
>>>>>
>>>>> Looking at the executor hadoop-w-2, I see the status is "Loading" .
>>>>> Why
>>>>> different statuses, and what does that mean?
>>>>>
>>>>> Please see below for details:
>>>>>
>>>>> ID: worker-20150202144112-hadoop-w-2.c.fi-mdd-poc.internal-58685
>>>>> Master URL: spark://hadoop-m:7077
>>>>> Cores: 16 (16 Used)
>>>>> Memory: 82.0 GB (82.0 GB Used)
>>>>> Back to Master
>>>>>
>>>>> Running Executors (1)
>>>>>
>>>>> ExecutorID      Cores   State   Memory  Job Details     Logs
>>>>> 0       16      LOADING 82.0 GB
>>>>> ID: app-20150202152154-0001
>>>>> Name: Simple File Merge Application
>>>>> User: hadoop
>>>>> stdout stderr
>>>>>
>>>>> Thank you,
>>>>>
>>>>> Ami
>>>>>
>>>>>
>>>>>
>>>>> --
>>>>> View this message in context:
>>>>> http://apache-spark-user-list.1001560.n3.nabble.com/Loading-status-tp21468.html
>>>>> Sent from the Apache Spark User List mailing list archive at
>>>>> Nabble.com.
>>>>>
>>>>> ---------------------------------------------------------------------
>>>>> To unsubscribe, e-mail: user-unsubscribe@spark.apache.org
>>>>> For additional commands, e-mail: user-help@spark.apache.org
>>>>>
>>>>>
>>>>
>>
>

Mime
View raw message