Subscribe

17 views
Skip to first unread message

william...@ubs.com

unread,
Sep 22, 2015, 9:39:46 AM9/22/15
to spark-not...@googlegroups.com

 

Please can I subscribe J

disclaim.txt

andy petrella

unread,
Sep 22, 2015, 11:52:32 PM9/22/15
to william...@ubs.com, spark-not...@googlegroups.com

you are, however you better come to gitter, it's easier and live ^^


Le mar. 22 sept. 2015 14:39,  <william...@ubs.com> a écrit :

 

Please can I subscribe J

--
You received this message because you are subscribed to the Google Groups "spark-notebook-user" group.
To unsubscribe from this group and stop receiving emails from it, send an email to spark-notebook-...@googlegroups.com.
To post to this group, send email to spark-not...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/spark-notebook-user/8D9222BC4E5F0C4BACFA55C11A9C129F2999C988%40NLDNC501PN3.UBSPROD.MSAD.UBS.NET.
For more options, visit https://groups.google.com/d/optout.
--
andy

william...@ubs.com

unread,
Oct 1, 2015, 12:48:48 PM10/1/15
to andy.p...@gmail.com, spark-not...@googlegroups.com

 

Couple of issues we're seeing (is this the right place to report these?):

 

-          If you change the spark master in the notebook metadata and then do a kernel reset, it doesn't actually change the master. I think it tries to, but the sc variable still points to the old master

-          When we run on spark cluster, sometime tasks just hang and on the worker error log we see this message

 

15/10/01 17:45:29 WARN ReliableDeliverySupervisor: Association with remote system [akka.tcp://spark...@139.149.52.141:56371] has failed, address is now gated for [5000] ms. Reason is: [Disassociated].

disclaim.txt

william...@ubs.com

unread,
Oct 1, 2015, 12:57:33 PM10/1/15
to andy.p...@gmail.com, spark-not...@googlegroups.com

 

And on the notebook log we see this:

2015-10-01 17:56:25,024  WARN [Timer-0] (org.apache.spark.scheduler.TaskSchedulerImpl) - Initial job has not accepted any resources; check your cluster UI to ensure that workers are registered and have sufficient resources

 

However if I run under the spark shell with the same settings the job executes fine

 

 

From: Rubens, William
Sent: 01 October 2015 17:49
To: 'andy petrella'; spark-not...@googlegroups.com
Subject: RE: Subscribe

 

 

Couple of issues we're seeing (is this the right place to report these?):

 

-          If you change the spark master in the notebook metadata and then do a kernel reset, it doesn't actually change the master. I think it tries to, but the sc variable still points to the old master

-          When we run on spark cluster, sometime tasks just hang and on the worker error log we see this message

 

15/10/01 17:45:29 WARN ReliableDeliverySupervisor: Association with remote system [akka.tcp://spark...@139.149.52.141:56371] has failed, address is now gated for [5000] ms. Reason is: [Disassociated].

 

 

From: andy petrella [mailto:andy.p...@gmail.com]
Sent: 23 September 2015 04:52
To: Rubens, William; spark-not...@googlegroups.com
Subject: Re: Subscribe

 

you are, however you better come to gitter, it's easier and live ^^

disclaim.txt
Reply all
Reply to author
Forward
0 new messages