You are viewing a plain text version of this content. The canonical link for it is here.
Posted to users@zeppelin.apache.org by Guilherme Silveira <gu...@gmail.com> on 2016/05/16 13:06:18 UTC
Zeppelin error
> Hi Folks,
>
> I am evaluating zeppelin right now but I am facing some annoying issues.
> I would like to hear from the community, maybe I am doing something
wrong.
>
> The number #1 annoying thing is unpredictable of some interpreters. Every
now and then I get a "pyspark interpleter not found". This is pretty
common, and it is not related to the source code itself. The same source
code, in different paragraphs, get different results. As you can see in the
example attached, the source code is exactly the same, and one paragraph is
right above the other. When I receive a pyspark interpreter not found,
there is nothing I can do besides remove the paragraph, insert a new one
and pray that the same error does not return. I would say that 30% of all
paragraphs get this error. How can I debug this?
>
>
>
>
> The second error is related to the spark configuration. Spark libraries
are working as expected and I can perform some basic ETL.
> However, I would like to generate the RDD from a jdbc connection, like
this:
>
> I thought that the only thing I needed to do is configure this:
>
> What else do I need?
>
> _______________________________.
RE: Zeppelin error
Posted by Anton Bubna-Litic <An...@quantium.com.au>.
This issue was raised earlier this year: https://issues.apache.org/jira/browse/ZEPPELIN-577
________________________________
linkedin.com/company/quantium<http://www.linkedin.com/company/quantium>
facebook.com/QuantiumAustralia<http://www.facebook.com/QuantiumAustralia>
twitter.com/QuantiumAU<http://www.twitter.com/QuantiumAU>
The contents of this email, including attachments, may be confidential information. If you are not the intended recipient, any use, disclosure or copying of the information is unauthorised. If you have received this email in error, we would be grateful if you would notify us immediately by email reply, phone (+ 61 2 9292 6400) or fax (+ 61 2 9292 6444) and delete the message from your system.
From: Guilherme Silveira [mailto:guilhermecgsspam@gmail.com]
Sent: Wednesday, 18 May 2016 12:44 AM
To: users@zeppelin.incubator.apache.org
Subject: Re: Zeppelin error
Looks like Chris's answer is right
ERROR:
"text":"%pyspark\r\n\r\ndef formatClusterName(index):\r\n return 'ClusterCenter_' + str(index +1) ",
SUCCESS:
"text":"%pyspark\n\ndef formatClusterName(index):\n return 'ClusterCenter_' + str(index + 1)",
Both look the same in the UI
Em 17 de mai de 2016 10:18, "Felix Cheung" <fe...@hotmail.com>> escreveu:
That's a great solution. If one of you don't mind opening a JIRA for this we should investigate and fix this lien ending check issue.
On Tue, May 17, 2016 at 6:05 AM -0700, "Chris Winne" <ch...@missoula.mine.nu>> wrote:
On Mon, May 16, 2016 at 10:06:18AM -0300, Guilherme Silveira wrote:
> > Hi Folks,
> >
> > I am evaluating zeppelin right now but I am facing some annoying issues.
> > I would like to hear from the community, maybe I am doing something
> wrong.
> >
> > The number #1 annoying thing is unpredictable of some interpreters.
> Every now and then I get a "pyspark interpleter not found". This is pretty
> common, and it is not related to the source code itself. The same source
> code, in different paragraphs, get different results. As you can see in
> the example attached, the source code is exactly the same, and one
> paragraph is right above the other. When I receive a pyspark interpreter
> not found, there is nothing I can do besides remove the paragraph, insert
> a new one and pray that the same error does not return. I would say that
> 30% of all paragraphs get this error. How can I debug this?
> >
> >
I have found that when code was pasted from an editor on windows
having evil (editorial..) CRLF endings, I had a similar problem (i.e.
%pyspark interpreter not found). Once I changed my editor to default to
LF endings, the pasting issue was gone. This can be checked by
examining the <notebook>.json file in the notebook directory of
Zeppelin. Look for a \r\n between code lines instead of simply a \n.
Chris
Re: Zeppelin error
Posted by Guilherme Silveira <gu...@gmail.com>.
Looks like Chris's answer is right
ERROR:
"text":"%pyspark\r\n\r\ndef formatClusterName(index):\r\n return
'ClusterCenter_' + str(index +1) ",
SUCCESS:
"text":"%pyspark\n\ndef formatClusterName(index):\n return
'ClusterCenter_' + str(index + 1)",
Both look the same in the UI
Em 17 de mai de 2016 10:18, "Felix Cheung" <fe...@hotmail.com>
escreveu:
> That's a great solution. If one of you don't mind opening a JIRA for this
> we should investigate and fix this lien ending check issue.
>
>
>
>
>
> On Tue, May 17, 2016 at 6:05 AM -0700, "Chris Winne" <
> chris@missoula.mine.nu> wrote:
>
> On Mon, May 16, 2016 at 10:06:18AM -0300, Guilherme Silveira wrote:
> > > Hi Folks,
> > >
> > > I am evaluating zeppelin right now but I am facing some annoying
> issues.
> > > I would like to hear from the community, maybe I am doing something
> > wrong.
> > >
> > > The number #1 annoying thing is unpredictable of some interpreters.
> > Every now and then I get a "pyspark interpleter not found". This is
> pretty
> > common, and it is not related to the source code itself. The same
> source
> > code, in different paragraphs, get different results. As you can see
> in
> > the example attached, the source code is exactly the same, and one
> > paragraph is right above the other. When I receive a pyspark
> interpreter
> > not found, there is nothing I can do besides remove the paragraph,
> insert
> > a new one and pray that the same error does not return. I would say
> that
> > 30% of all paragraphs get this error. How can I debug this?
> > >
> > >
>
> I have found that when code was pasted from an editor on windows
> having evil (editorial..) CRLF endings, I had a similar problem (i.e.
> %pyspark interpreter not found). Once I changed my editor to default to
> LF endings, the pasting issue was gone. This can be checked by
> examining the <notebook>.json file in the notebook directory of
> Zeppelin. Look for a \r\n between code lines instead of simply a \n.
>
> Chris
>
Re: Zeppelin error
Posted by Felix Cheung <fe...@hotmail.com>.
That's a great solution. If one of you don't mind opening a JIRA for this we should investigate and fix this lien ending check issue.
On Tue, May 17, 2016 at 6:05 AM -0700, "Chris Winne" <ch...@missoula.mine.nu> wrote:
On Mon, May 16, 2016 at 10:06:18AM -0300, Guilherme Silveira wrote:
> > Hi Folks,
> >
> > I am evaluating zeppelin right now but I am facing some annoying issues.
> > I would like to hear from the community, maybe I am doing something
> wrong.
> >
> > The number #1 annoying thing is unpredictable of some interpreters.
> Every now and then I get a "pyspark interpleter not found". This is pretty
> common, and it is not related to the source code itself. The same source
> code, in different paragraphs, get different results. As you can see in
> the example attached, the source code is exactly the same, and one
> paragraph is right above the other. When I receive a pyspark interpreter
> not found, there is nothing I can do besides remove the paragraph, insert
> a new one and pray that the same error does not return. I would say that
> 30% of all paragraphs get this error. How can I debug this?
> >
> >
I have found that when code was pasted from an editor on windows
having evil (editorial..) CRLF endings, I had a similar problem (i.e.
%pyspark interpreter not found). Once I changed my editor to default to
LF endings, the pasting issue was gone. This can be checked by
examining the <notebook>.json file in the notebook directory of
Zeppelin. Look for a \r\n between code lines instead of simply a \n.
Chris
Re: Zeppelin error
Posted by Chris Winne <ch...@missoula.mine.nu>.
On Mon, May 16, 2016 at 10:06:18AM -0300, Guilherme Silveira wrote:
> > Hi Folks,
> >
> > I am evaluating zeppelin right now but I am facing some annoying issues.
> > I would like to hear from the community, maybe I am doing something
> wrong.
> >
> > The number #1 annoying thing is unpredictable of some interpreters.
> Every now and then I get a "pyspark interpleter not found". This is pretty
> common, and it is not related to the source code itself. The same source
> code, in different paragraphs, get different results. As you can see in
> the example attached, the source code is exactly the same, and one
> paragraph is right above the other. When I receive a pyspark interpreter
> not found, there is nothing I can do besides remove the paragraph, insert
> a new one and pray that the same error does not return. I would say that
> 30% of all paragraphs get this error. How can I debug this?
> >
> >
I have found that when code was pasted from an editor on windows
having evil (editorial..) CRLF endings, I had a similar problem (i.e.
%pyspark interpreter not found). Once I changed my editor to default to
LF endings, the pasting issue was gone. This can be checked by
examining the <notebook>.json file in the notebook directory of
Zeppelin. Look for a \r\n between code lines instead of simply a \n.
Chris
Re: Zeppelin error
Posted by moon soo Lee <mo...@apache.org>.
Hi,
Really appreciate for sharing the problem.
This is how Zeppelin extracting interpreter name
https://github.com/apache/incubator-zeppelin/blob/master/zeppelin-zengine/src/main/java/org/apache/zeppelin/notebook/Paragraph.java#L137
This is how Interpreter is selected from the extracted name
https://github.com/apache/incubator-zeppelin/blob/master/zeppelin-zengine/src/main/java/org/apache/zeppelin/notebook/NoteInterpreterLoader.java#L110
If that problem happens, could you some log messages into those two those
methods and see where the problems come from?
Thanks,
moon
On Mon, May 16, 2016 at 6:06 AM Guilherme Silveira <
guilhermecgsspam@gmail.com> wrote:
>
> > Hi Folks,
> >
> > I am evaluating zeppelin right now but I am facing some annoying issues.
> > I would like to hear from the community, maybe I am doing something
> wrong.
> >
> > The number #1 annoying thing is unpredictable of some interpreters.
> Every now and then I get a "pyspark interpleter not found". This is pretty
> common, and it is not related to the source code itself. The same source
> code, in different paragraphs, get different results. As you can see in the
> example attached, the source code is exactly the same, and one paragraph is
> right above the other. When I receive a pyspark interpreter not found,
> there is nothing I can do besides remove the paragraph, insert a new one
> and pray that the same error does not return. I would say that 30% of all
> paragraphs get this error. How can I debug this?
> >
> >
> >
> >
> > The second error is related to the spark configuration. Spark libraries
> are working as expected and I can perform some basic ETL.
> > However, I would like to generate the RDD from a jdbc connection, like
> this:
> >
> > I thought that the only thing I needed to do is configure this:
> >
> > What else do I need?
> >
> > _______________________________.
>