You are viewing a plain text version of this content. The canonical link for it is here.
Posted to common-user@hadoop.apache.org by Tom Deutsch <td...@us.ibm.com> on 2012/09/20 01:39:17 UTC
Re: IBM big insights distribution
Serge - let's take this off list. Ping me directly and we'll take it from there.
---------------------------------------
Sent from my Blackberry so please excuse typing and spelling errors.
----- Original Message -----
From: Serge Blazhiyevskyy [Serge.Blazhiyevskyy@nice.com]
Sent: 09/19/2012 03:16 PM CST
To: "user@hadoop.apache.org" <us...@hadoop.apache.org>
Subject: IBM big insights distribution
Hi all,
Anybody had a success compiling Map reduce jobs with big insights distro on MAC?
Seems like it requires IBM java JDK and it might not be available on MAC.
Is there a way work around it?
Thanks
Serge
Re: IBM big insights distribution
Posted by Michael Segel <mi...@hotmail.com>.
I was looking at it from outside of the support issue, but from an industry related topic.
I mean, lets talk about Solaris support. I believe that its still there in the Apache release while none of the commercial distros support it.
To your point, I can see how it could confuse some people...
On Sep 24, 2012, at 5:58 PM, Arun C Murthy <ac...@hortonworks.com> wrote:
>
> On Sep 20, 2012, at 11:26 AM, Andy Isaacson wrote:
>
>> If, OTOH, someone is experiencing a problem that is probably down to
>> core Hadoop and just happens to be using a particular distro, I think
>> user@hadoop is a perfectly reasonable place to discuss that, and
>> there's no particular need to hide the fact that a specific distro is
>> in use, it's not as if The Name Cannot Be Spoken or something.
>
> That has it's own shades of gray - different vendors may choose to include different patch-sets in their distro, even for 'core Hadoop'.
>
> In such cases, you get issues/bug-reports which are, at best, confusing to users of Apache Hadoop... etc.
>
> As Andrew mentioned, it seems simpler to keep things on Apache lists limited to Apache Hadoop.
>
> Arun
Re: IBM big insights distribution
Posted by Michael Segel <mi...@hotmail.com>.
I was looking at it from outside of the support issue, but from an industry related topic.
I mean, lets talk about Solaris support. I believe that its still there in the Apache release while none of the commercial distros support it.
To your point, I can see how it could confuse some people...
On Sep 24, 2012, at 5:58 PM, Arun C Murthy <ac...@hortonworks.com> wrote:
>
> On Sep 20, 2012, at 11:26 AM, Andy Isaacson wrote:
>
>> If, OTOH, someone is experiencing a problem that is probably down to
>> core Hadoop and just happens to be using a particular distro, I think
>> user@hadoop is a perfectly reasonable place to discuss that, and
>> there's no particular need to hide the fact that a specific distro is
>> in use, it's not as if The Name Cannot Be Spoken or something.
>
> That has it's own shades of gray - different vendors may choose to include different patch-sets in their distro, even for 'core Hadoop'.
>
> In such cases, you get issues/bug-reports which are, at best, confusing to users of Apache Hadoop... etc.
>
> As Andrew mentioned, it seems simpler to keep things on Apache lists limited to Apache Hadoop.
>
> Arun
Re: IBM big insights distribution
Posted by Michael Segel <mi...@hotmail.com>.
I was looking at it from outside of the support issue, but from an industry related topic.
I mean, lets talk about Solaris support. I believe that its still there in the Apache release while none of the commercial distros support it.
To your point, I can see how it could confuse some people...
On Sep 24, 2012, at 5:58 PM, Arun C Murthy <ac...@hortonworks.com> wrote:
>
> On Sep 20, 2012, at 11:26 AM, Andy Isaacson wrote:
>
>> If, OTOH, someone is experiencing a problem that is probably down to
>> core Hadoop and just happens to be using a particular distro, I think
>> user@hadoop is a perfectly reasonable place to discuss that, and
>> there's no particular need to hide the fact that a specific distro is
>> in use, it's not as if The Name Cannot Be Spoken or something.
>
> That has it's own shades of gray - different vendors may choose to include different patch-sets in their distro, even for 'core Hadoop'.
>
> In such cases, you get issues/bug-reports which are, at best, confusing to users of Apache Hadoop... etc.
>
> As Andrew mentioned, it seems simpler to keep things on Apache lists limited to Apache Hadoop.
>
> Arun
Re: IBM big insights distribution
Posted by Michael Segel <mi...@hotmail.com>.
I was looking at it from outside of the support issue, but from an industry related topic.
I mean, lets talk about Solaris support. I believe that its still there in the Apache release while none of the commercial distros support it.
To your point, I can see how it could confuse some people...
On Sep 24, 2012, at 5:58 PM, Arun C Murthy <ac...@hortonworks.com> wrote:
>
> On Sep 20, 2012, at 11:26 AM, Andy Isaacson wrote:
>
>> If, OTOH, someone is experiencing a problem that is probably down to
>> core Hadoop and just happens to be using a particular distro, I think
>> user@hadoop is a perfectly reasonable place to discuss that, and
>> there's no particular need to hide the fact that a specific distro is
>> in use, it's not as if The Name Cannot Be Spoken or something.
>
> That has it's own shades of gray - different vendors may choose to include different patch-sets in their distro, even for 'core Hadoop'.
>
> In such cases, you get issues/bug-reports which are, at best, confusing to users of Apache Hadoop... etc.
>
> As Andrew mentioned, it seems simpler to keep things on Apache lists limited to Apache Hadoop.
>
> Arun
Re: IBM big insights distribution
Posted by Arun C Murthy <ac...@hortonworks.com>.
On Sep 20, 2012, at 11:26 AM, Andy Isaacson wrote:
> If, OTOH, someone is experiencing a problem that is probably down to
> core Hadoop and just happens to be using a particular distro, I think
> user@hadoop is a perfectly reasonable place to discuss that, and
> there's no particular need to hide the fact that a specific distro is
> in use, it's not as if The Name Cannot Be Spoken or something.
That has it's own shades of gray - different vendors may choose to include different patch-sets in their distro, even for 'core Hadoop'.
In such cases, you get issues/bug-reports which are, at best, confusing to users of Apache Hadoop... etc.
As Andrew mentioned, it seems simpler to keep things on Apache lists limited to Apache Hadoop.
Arun
Re: IBM big insights distribution
Posted by Arun C Murthy <ac...@hortonworks.com>.
On Sep 20, 2012, at 11:26 AM, Andy Isaacson wrote:
> If, OTOH, someone is experiencing a problem that is probably down to
> core Hadoop and just happens to be using a particular distro, I think
> user@hadoop is a perfectly reasonable place to discuss that, and
> there's no particular need to hide the fact that a specific distro is
> in use, it's not as if The Name Cannot Be Spoken or something.
That has it's own shades of gray - different vendors may choose to include different patch-sets in their distro, even for 'core Hadoop'.
In such cases, you get issues/bug-reports which are, at best, confusing to users of Apache Hadoop... etc.
As Andrew mentioned, it seems simpler to keep things on Apache lists limited to Apache Hadoop.
Arun
Re: IBM big insights distribution
Posted by Arun C Murthy <ac...@hortonworks.com>.
On Sep 20, 2012, at 11:26 AM, Andy Isaacson wrote:
> If, OTOH, someone is experiencing a problem that is probably down to
> core Hadoop and just happens to be using a particular distro, I think
> user@hadoop is a perfectly reasonable place to discuss that, and
> there's no particular need to hide the fact that a specific distro is
> in use, it's not as if The Name Cannot Be Spoken or something.
That has it's own shades of gray - different vendors may choose to include different patch-sets in their distro, even for 'core Hadoop'.
In such cases, you get issues/bug-reports which are, at best, confusing to users of Apache Hadoop... etc.
As Andrew mentioned, it seems simpler to keep things on Apache lists limited to Apache Hadoop.
Arun
Re: IBM big insights distribution
Posted by Arun C Murthy <ac...@hortonworks.com>.
On Sep 20, 2012, at 11:26 AM, Andy Isaacson wrote:
> If, OTOH, someone is experiencing a problem that is probably down to
> core Hadoop and just happens to be using a particular distro, I think
> user@hadoop is a perfectly reasonable place to discuss that, and
> there's no particular need to hide the fact that a specific distro is
> in use, it's not as if The Name Cannot Be Spoken or something.
That has it's own shades of gray - different vendors may choose to include different patch-sets in their distro, even for 'core Hadoop'.
In such cases, you get issues/bug-reports which are, at best, confusing to users of Apache Hadoop... etc.
As Andrew mentioned, it seems simpler to keep things on Apache lists limited to Apache Hadoop.
Arun
Re: IBM big insights distribution
Posted by Andy Isaacson <ad...@cloudera.com>.
On Thu, Sep 20, 2012 at 5:24 AM, Michael Segel
<mi...@hotmail.com> wrote:
> Why is it that when anyone asks a question about IBM Tom wants to take it off line?
To be fair, most vendors tend to redirect distro-specific discussion
to non-apache.org forums. Cloudera has a cdh-user list, MapR has web
forums and a support address IIRC, Hortonworks has community forums,
et cetera. I have no problem recognizing that an apache-hosted list is
not the right place to discuss a vendor-specific problem.
If, OTOH, someone is experiencing a problem that is probably down to
core Hadoop and just happens to be using a particular distro, I think
user@hadoop is a perfectly reasonable place to discuss that, and
there's no particular need to hide the fact that a specific distro is
in use, it's not as if The Name Cannot Be Spoken or something.
> Not very social.
Different user communities have different norms for behavior, and make
different technical choices for communication medium based on that.
Seems pretty reasonable to me. I may find some people's customs odd,
but I also recognize they probably feel the same way about me. :) If I
were Tom I'd redirect to a different public forum rather than
attempting to help each user one-on-one, but I'm not about to tell him
how to organize his life.
-andy
Re: IBM big insights distribution
Posted by Andy Isaacson <ad...@cloudera.com>.
On Thu, Sep 20, 2012 at 5:24 AM, Michael Segel
<mi...@hotmail.com> wrote:
> Why is it that when anyone asks a question about IBM Tom wants to take it off line?
To be fair, most vendors tend to redirect distro-specific discussion
to non-apache.org forums. Cloudera has a cdh-user list, MapR has web
forums and a support address IIRC, Hortonworks has community forums,
et cetera. I have no problem recognizing that an apache-hosted list is
not the right place to discuss a vendor-specific problem.
If, OTOH, someone is experiencing a problem that is probably down to
core Hadoop and just happens to be using a particular distro, I think
user@hadoop is a perfectly reasonable place to discuss that, and
there's no particular need to hide the fact that a specific distro is
in use, it's not as if The Name Cannot Be Spoken or something.
> Not very social.
Different user communities have different norms for behavior, and make
different technical choices for communication medium based on that.
Seems pretty reasonable to me. I may find some people's customs odd,
but I also recognize they probably feel the same way about me. :) If I
were Tom I'd redirect to a different public forum rather than
attempting to help each user one-on-one, but I'm not about to tell him
how to organize his life.
-andy
Re: IBM big insights distribution
Posted by John McPherson <jm...@us.ibm.com>.
Just clarifying because of the reference to mainframe....
IBM BigInsights runs on clusters of x86 servers.
John
From: Michael Segel <mi...@hotmail.com>
To: user@hadoop.apache.org
Cc: Serge.Blazhiyevskyy@nice.com
Date: 09/20/2012 05:25 AM
Subject: Re: IBM big insights distribution
Why is it that when anyone asks a question about IBM Tom wants to take it
off line?
Not very social.
I personally find it interesting that IBM requires that their products are
tied to IBM's JVM.
I also wonder about the cost of running Hadoop on the mainframe.
(Linux LPARs while cheaper, still cost mips. ) [Forgive me, its been a
while since I spoke mainframe]
So what's the real advantage from an architecture point of view?
On Sep 19, 2012, at 6:39 PM, Tom Deutsch <td...@us.ibm.com> wrote:
> Serge - let's take this off list. Ping me directly and we'll take it
from there.
>
> ---------------------------------------
> Sent from my Blackberry so please excuse typing and spelling errors.
>
>
> ----- Original Message -----
> From: Serge Blazhiyevskyy [Serge.Blazhiyevskyy@nice.com]
> Sent: 09/19/2012 03:16 PM CST
> To: "user@hadoop.apache.org" <us...@hadoop.apache.org>
> Subject: IBM big insights distribution
>
>
>
> Hi all,
>
> Anybody had a success compiling Map reduce jobs with big insights distro
on MAC?
>
> Seems like it requires IBM java JDK and it might not be available on
MAC.
>
> Is there a way work around it?
>
>
> Thanks
> Serge
>
>
>
>
Re: IBM big insights distribution
Posted by Andrew Purtell <ap...@apache.org>.
As a community observer, I see some tension between vendor marketing
efforts that *some* may consider to be disparaging of Apache community
efforts and work product (for example the HDFS file system); and good
faith efforts from the same vendors to participate in Apache community
efforts and work product, in other areas. That tension seems
unavoidable to me and is something vendors and the community will come
to terms with over time. An easy way to minimize potential conflict,
however, is to redirect vendor specific discussion to vendor mailing
lists and forums. Seems a no-brainer.
On Thu, Sep 20, 2012 at 5:49 AM, Ted Dunning <td...@maprtech.com> wrote:
>
>
> On Thu, Sep 20, 2012 at 5:24 AM, Michael Segel <mi...@hotmail.com>
> wrote:
>>
>> Why is it that when anyone asks a question about IBM Tom wants to take it
>> off line?
>>
>> Not very social.
>
>
> Well, giving the mesquite broiling that anyone gets if they even mention any
> product competitive with certain Hadoop distributions, I tend to do that as
> well. I have even gotten private threatening emails telling me not to
> attend public Apache events.
>
> Not social. Check. Doesn't help the group find out about alternatives.
> Check.
>
> But it lowers the temperature on my side of the keyboard to not mention MapR
> in public. I imagine that Tom feels much the same way.
--
Best regards,
- Andy
Problems worthy of attack prove their worth by hitting back. - Piet
Hein (via Tom White)
Re: IBM big insights distribution
Posted by Tom Deutsch <td...@us.ibm.com>.
Exactly Ted, and just trying to be respectful of this being an Apache
mailing list.
Snarky behavior devalues everyone's efforts, so trying to head that off as
it doesn't have any place on a community list or forum.
From: Ted Dunning <td...@maprtech.com>
To: user@hadoop.apache.org,
Cc: Serge.Blazhiyevskyy@nice.com
Date: 09/20/2012 05:50 AM
Subject: Re: IBM big insights distribution
On Thu, Sep 20, 2012 at 5:24 AM, Michael Segel <mi...@hotmail.com>
wrote:
Why is it that when anyone asks a question about IBM Tom wants to take it
off line?
Not very social.
Well, giving the mesquite broiling that anyone gets if they even mention
any product competitive with certain Hadoop distributions, I tend to do
that as well. I have even gotten private threatening emails telling me not
to attend public Apache events.
Not social. Check. Doesn't help the group find out about alternatives.
Check.
But it lowers the temperature on my side of the keyboard to not mention
MapR in public. I imagine that Tom feels much the same way.
Re: IBM big insights distribution
Posted by Tom Deutsch <td...@us.ibm.com>.
Exactly Ted, and just trying to be respectful of this being an Apache
mailing list.
Snarky behavior devalues everyone's efforts, so trying to head that off as
it doesn't have any place on a community list or forum.
From: Ted Dunning <td...@maprtech.com>
To: user@hadoop.apache.org,
Cc: Serge.Blazhiyevskyy@nice.com
Date: 09/20/2012 05:50 AM
Subject: Re: IBM big insights distribution
On Thu, Sep 20, 2012 at 5:24 AM, Michael Segel <mi...@hotmail.com>
wrote:
Why is it that when anyone asks a question about IBM Tom wants to take it
off line?
Not very social.
Well, giving the mesquite broiling that anyone gets if they even mention
any product competitive with certain Hadoop distributions, I tend to do
that as well. I have even gotten private threatening emails telling me not
to attend public Apache events.
Not social. Check. Doesn't help the group find out about alternatives.
Check.
But it lowers the temperature on my side of the keyboard to not mention
MapR in public. I imagine that Tom feels much the same way.
Re: IBM big insights distribution
Posted by Tom Deutsch <td...@us.ibm.com>.
Exactly Ted, and just trying to be respectful of this being an Apache
mailing list.
Snarky behavior devalues everyone's efforts, so trying to head that off as
it doesn't have any place on a community list or forum.
From: Ted Dunning <td...@maprtech.com>
To: user@hadoop.apache.org,
Cc: Serge.Blazhiyevskyy@nice.com
Date: 09/20/2012 05:50 AM
Subject: Re: IBM big insights distribution
On Thu, Sep 20, 2012 at 5:24 AM, Michael Segel <mi...@hotmail.com>
wrote:
Why is it that when anyone asks a question about IBM Tom wants to take it
off line?
Not very social.
Well, giving the mesquite broiling that anyone gets if they even mention
any product competitive with certain Hadoop distributions, I tend to do
that as well. I have even gotten private threatening emails telling me not
to attend public Apache events.
Not social. Check. Doesn't help the group find out about alternatives.
Check.
But it lowers the temperature on my side of the keyboard to not mention
MapR in public. I imagine that Tom feels much the same way.
Re: IBM big insights distribution
Posted by Andrew Purtell <ap...@apache.org>.
As a community observer, I see some tension between vendor marketing
efforts that *some* may consider to be disparaging of Apache community
efforts and work product (for example the HDFS file system); and good
faith efforts from the same vendors to participate in Apache community
efforts and work product, in other areas. That tension seems
unavoidable to me and is something vendors and the community will come
to terms with over time. An easy way to minimize potential conflict,
however, is to redirect vendor specific discussion to vendor mailing
lists and forums. Seems a no-brainer.
On Thu, Sep 20, 2012 at 5:49 AM, Ted Dunning <td...@maprtech.com> wrote:
>
>
> On Thu, Sep 20, 2012 at 5:24 AM, Michael Segel <mi...@hotmail.com>
> wrote:
>>
>> Why is it that when anyone asks a question about IBM Tom wants to take it
>> off line?
>>
>> Not very social.
>
>
> Well, giving the mesquite broiling that anyone gets if they even mention any
> product competitive with certain Hadoop distributions, I tend to do that as
> well. I have even gotten private threatening emails telling me not to
> attend public Apache events.
>
> Not social. Check. Doesn't help the group find out about alternatives.
> Check.
>
> But it lowers the temperature on my side of the keyboard to not mention MapR
> in public. I imagine that Tom feels much the same way.
--
Best regards,
- Andy
Problems worthy of attack prove their worth by hitting back. - Piet
Hein (via Tom White)
Re: IBM big insights distribution
Posted by Tom Deutsch <td...@us.ibm.com>.
Exactly Ted, and just trying to be respectful of this being an Apache
mailing list.
Snarky behavior devalues everyone's efforts, so trying to head that off as
it doesn't have any place on a community list or forum.
From: Ted Dunning <td...@maprtech.com>
To: user@hadoop.apache.org,
Cc: Serge.Blazhiyevskyy@nice.com
Date: 09/20/2012 05:50 AM
Subject: Re: IBM big insights distribution
On Thu, Sep 20, 2012 at 5:24 AM, Michael Segel <mi...@hotmail.com>
wrote:
Why is it that when anyone asks a question about IBM Tom wants to take it
off line?
Not very social.
Well, giving the mesquite broiling that anyone gets if they even mention
any product competitive with certain Hadoop distributions, I tend to do
that as well. I have even gotten private threatening emails telling me not
to attend public Apache events.
Not social. Check. Doesn't help the group find out about alternatives.
Check.
But it lowers the temperature on my side of the keyboard to not mention
MapR in public. I imagine that Tom feels much the same way.
Re: IBM big insights distribution
Posted by Andrew Purtell <ap...@apache.org>.
As a community observer, I see some tension between vendor marketing
efforts that *some* may consider to be disparaging of Apache community
efforts and work product (for example the HDFS file system); and good
faith efforts from the same vendors to participate in Apache community
efforts and work product, in other areas. That tension seems
unavoidable to me and is something vendors and the community will come
to terms with over time. An easy way to minimize potential conflict,
however, is to redirect vendor specific discussion to vendor mailing
lists and forums. Seems a no-brainer.
On Thu, Sep 20, 2012 at 5:49 AM, Ted Dunning <td...@maprtech.com> wrote:
>
>
> On Thu, Sep 20, 2012 at 5:24 AM, Michael Segel <mi...@hotmail.com>
> wrote:
>>
>> Why is it that when anyone asks a question about IBM Tom wants to take it
>> off line?
>>
>> Not very social.
>
>
> Well, giving the mesquite broiling that anyone gets if they even mention any
> product competitive with certain Hadoop distributions, I tend to do that as
> well. I have even gotten private threatening emails telling me not to
> attend public Apache events.
>
> Not social. Check. Doesn't help the group find out about alternatives.
> Check.
>
> But it lowers the temperature on my side of the keyboard to not mention MapR
> in public. I imagine that Tom feels much the same way.
--
Best regards,
- Andy
Problems worthy of attack prove their worth by hitting back. - Piet
Hein (via Tom White)
Re: IBM big insights distribution
Posted by Andrew Purtell <ap...@apache.org>.
As a community observer, I see some tension between vendor marketing
efforts that *some* may consider to be disparaging of Apache community
efforts and work product (for example the HDFS file system); and good
faith efforts from the same vendors to participate in Apache community
efforts and work product, in other areas. That tension seems
unavoidable to me and is something vendors and the community will come
to terms with over time. An easy way to minimize potential conflict,
however, is to redirect vendor specific discussion to vendor mailing
lists and forums. Seems a no-brainer.
On Thu, Sep 20, 2012 at 5:49 AM, Ted Dunning <td...@maprtech.com> wrote:
>
>
> On Thu, Sep 20, 2012 at 5:24 AM, Michael Segel <mi...@hotmail.com>
> wrote:
>>
>> Why is it that when anyone asks a question about IBM Tom wants to take it
>> off line?
>>
>> Not very social.
>
>
> Well, giving the mesquite broiling that anyone gets if they even mention any
> product competitive with certain Hadoop distributions, I tend to do that as
> well. I have even gotten private threatening emails telling me not to
> attend public Apache events.
>
> Not social. Check. Doesn't help the group find out about alternatives.
> Check.
>
> But it lowers the temperature on my side of the keyboard to not mention MapR
> in public. I imagine that Tom feels much the same way.
--
Best regards,
- Andy
Problems worthy of attack prove their worth by hitting back. - Piet
Hein (via Tom White)
Re: IBM big insights distribution
Posted by Ted Dunning <td...@maprtech.com>.
On Thu, Sep 20, 2012 at 5:24 AM, Michael Segel <mi...@hotmail.com>wrote:
> Why is it that when anyone asks a question about IBM Tom wants to take it
> off line?
>
> Not very social.
>
Well, giving the mesquite broiling that anyone gets if they even mention
any product competitive with certain Hadoop distributions, I tend to do
that as well. I have even gotten private threatening emails telling me not
to attend public Apache events.
Not social. Check. Doesn't help the group find out about alternatives.
Check.
But it lowers the temperature on my side of the keyboard to not mention
MapR in public. I imagine that Tom feels much the same way.
Re: IBM big insights distribution
Posted by Ted Dunning <td...@maprtech.com>.
On Thu, Sep 20, 2012 at 5:24 AM, Michael Segel <mi...@hotmail.com>wrote:
> Why is it that when anyone asks a question about IBM Tom wants to take it
> off line?
>
> Not very social.
>
Well, giving the mesquite broiling that anyone gets if they even mention
any product competitive with certain Hadoop distributions, I tend to do
that as well. I have even gotten private threatening emails telling me not
to attend public Apache events.
Not social. Check. Doesn't help the group find out about alternatives.
Check.
But it lowers the temperature on my side of the keyboard to not mention
MapR in public. I imagine that Tom feels much the same way.
Re: IBM big insights distribution
Posted by John McPherson <jm...@us.ibm.com>.
Just clarifying because of the reference to mainframe....
IBM BigInsights runs on clusters of x86 servers.
John
From: Michael Segel <mi...@hotmail.com>
To: user@hadoop.apache.org
Cc: Serge.Blazhiyevskyy@nice.com
Date: 09/20/2012 05:25 AM
Subject: Re: IBM big insights distribution
Why is it that when anyone asks a question about IBM Tom wants to take it
off line?
Not very social.
I personally find it interesting that IBM requires that their products are
tied to IBM's JVM.
I also wonder about the cost of running Hadoop on the mainframe.
(Linux LPARs while cheaper, still cost mips. ) [Forgive me, its been a
while since I spoke mainframe]
So what's the real advantage from an architecture point of view?
On Sep 19, 2012, at 6:39 PM, Tom Deutsch <td...@us.ibm.com> wrote:
> Serge - let's take this off list. Ping me directly and we'll take it
from there.
>
> ---------------------------------------
> Sent from my Blackberry so please excuse typing and spelling errors.
>
>
> ----- Original Message -----
> From: Serge Blazhiyevskyy [Serge.Blazhiyevskyy@nice.com]
> Sent: 09/19/2012 03:16 PM CST
> To: "user@hadoop.apache.org" <us...@hadoop.apache.org>
> Subject: IBM big insights distribution
>
>
>
> Hi all,
>
> Anybody had a success compiling Map reduce jobs with big insights distro
on MAC?
>
> Seems like it requires IBM java JDK and it might not be available on
MAC.
>
> Is there a way work around it?
>
>
> Thanks
> Serge
>
>
>
>
Re: IBM big insights distribution
Posted by John McPherson <jm...@us.ibm.com>.
Just clarifying because of the reference to mainframe....
IBM BigInsights runs on clusters of x86 servers.
John
From: Michael Segel <mi...@hotmail.com>
To: user@hadoop.apache.org
Cc: Serge.Blazhiyevskyy@nice.com
Date: 09/20/2012 05:25 AM
Subject: Re: IBM big insights distribution
Why is it that when anyone asks a question about IBM Tom wants to take it
off line?
Not very social.
I personally find it interesting that IBM requires that their products are
tied to IBM's JVM.
I also wonder about the cost of running Hadoop on the mainframe.
(Linux LPARs while cheaper, still cost mips. ) [Forgive me, its been a
while since I spoke mainframe]
So what's the real advantage from an architecture point of view?
On Sep 19, 2012, at 6:39 PM, Tom Deutsch <td...@us.ibm.com> wrote:
> Serge - let's take this off list. Ping me directly and we'll take it
from there.
>
> ---------------------------------------
> Sent from my Blackberry so please excuse typing and spelling errors.
>
>
> ----- Original Message -----
> From: Serge Blazhiyevskyy [Serge.Blazhiyevskyy@nice.com]
> Sent: 09/19/2012 03:16 PM CST
> To: "user@hadoop.apache.org" <us...@hadoop.apache.org>
> Subject: IBM big insights distribution
>
>
>
> Hi all,
>
> Anybody had a success compiling Map reduce jobs with big insights distro
on MAC?
>
> Seems like it requires IBM java JDK and it might not be available on
MAC.
>
> Is there a way work around it?
>
>
> Thanks
> Serge
>
>
>
>
Re: IBM big insights distribution
Posted by John McPherson <jm...@us.ibm.com>.
Just clarifying because of the reference to mainframe....
IBM BigInsights runs on clusters of x86 servers.
John
From: Michael Segel <mi...@hotmail.com>
To: user@hadoop.apache.org
Cc: Serge.Blazhiyevskyy@nice.com
Date: 09/20/2012 05:25 AM
Subject: Re: IBM big insights distribution
Why is it that when anyone asks a question about IBM Tom wants to take it
off line?
Not very social.
I personally find it interesting that IBM requires that their products are
tied to IBM's JVM.
I also wonder about the cost of running Hadoop on the mainframe.
(Linux LPARs while cheaper, still cost mips. ) [Forgive me, its been a
while since I spoke mainframe]
So what's the real advantage from an architecture point of view?
On Sep 19, 2012, at 6:39 PM, Tom Deutsch <td...@us.ibm.com> wrote:
> Serge - let's take this off list. Ping me directly and we'll take it
from there.
>
> ---------------------------------------
> Sent from my Blackberry so please excuse typing and spelling errors.
>
>
> ----- Original Message -----
> From: Serge Blazhiyevskyy [Serge.Blazhiyevskyy@nice.com]
> Sent: 09/19/2012 03:16 PM CST
> To: "user@hadoop.apache.org" <us...@hadoop.apache.org>
> Subject: IBM big insights distribution
>
>
>
> Hi all,
>
> Anybody had a success compiling Map reduce jobs with big insights distro
on MAC?
>
> Seems like it requires IBM java JDK and it might not be available on
MAC.
>
> Is there a way work around it?
>
>
> Thanks
> Serge
>
>
>
>
Re: IBM big insights distribution
Posted by Ted Dunning <td...@maprtech.com>.
On Thu, Sep 20, 2012 at 5:24 AM, Michael Segel <mi...@hotmail.com>wrote:
> Why is it that when anyone asks a question about IBM Tom wants to take it
> off line?
>
> Not very social.
>
Well, giving the mesquite broiling that anyone gets if they even mention
any product competitive with certain Hadoop distributions, I tend to do
that as well. I have even gotten private threatening emails telling me not
to attend public Apache events.
Not social. Check. Doesn't help the group find out about alternatives.
Check.
But it lowers the temperature on my side of the keyboard to not mention
MapR in public. I imagine that Tom feels much the same way.
Re: IBM big insights distribution
Posted by Andy Isaacson <ad...@cloudera.com>.
On Thu, Sep 20, 2012 at 5:24 AM, Michael Segel
<mi...@hotmail.com> wrote:
> Why is it that when anyone asks a question about IBM Tom wants to take it off line?
To be fair, most vendors tend to redirect distro-specific discussion
to non-apache.org forums. Cloudera has a cdh-user list, MapR has web
forums and a support address IIRC, Hortonworks has community forums,
et cetera. I have no problem recognizing that an apache-hosted list is
not the right place to discuss a vendor-specific problem.
If, OTOH, someone is experiencing a problem that is probably down to
core Hadoop and just happens to be using a particular distro, I think
user@hadoop is a perfectly reasonable place to discuss that, and
there's no particular need to hide the fact that a specific distro is
in use, it's not as if The Name Cannot Be Spoken or something.
> Not very social.
Different user communities have different norms for behavior, and make
different technical choices for communication medium based on that.
Seems pretty reasonable to me. I may find some people's customs odd,
but I also recognize they probably feel the same way about me. :) If I
were Tom I'd redirect to a different public forum rather than
attempting to help each user one-on-one, but I'm not about to tell him
how to organize his life.
-andy
Re: IBM big insights distribution
Posted by Andy Isaacson <ad...@cloudera.com>.
On Thu, Sep 20, 2012 at 5:24 AM, Michael Segel
<mi...@hotmail.com> wrote:
> Why is it that when anyone asks a question about IBM Tom wants to take it off line?
To be fair, most vendors tend to redirect distro-specific discussion
to non-apache.org forums. Cloudera has a cdh-user list, MapR has web
forums and a support address IIRC, Hortonworks has community forums,
et cetera. I have no problem recognizing that an apache-hosted list is
not the right place to discuss a vendor-specific problem.
If, OTOH, someone is experiencing a problem that is probably down to
core Hadoop and just happens to be using a particular distro, I think
user@hadoop is a perfectly reasonable place to discuss that, and
there's no particular need to hide the fact that a specific distro is
in use, it's not as if The Name Cannot Be Spoken or something.
> Not very social.
Different user communities have different norms for behavior, and make
different technical choices for communication medium based on that.
Seems pretty reasonable to me. I may find some people's customs odd,
but I also recognize they probably feel the same way about me. :) If I
were Tom I'd redirect to a different public forum rather than
attempting to help each user one-on-one, but I'm not about to tell him
how to organize his life.
-andy
Re: IBM big insights distribution
Posted by Ted Dunning <td...@maprtech.com>.
On Thu, Sep 20, 2012 at 5:24 AM, Michael Segel <mi...@hotmail.com>wrote:
> Why is it that when anyone asks a question about IBM Tom wants to take it
> off line?
>
> Not very social.
>
Well, giving the mesquite broiling that anyone gets if they even mention
any product competitive with certain Hadoop distributions, I tend to do
that as well. I have even gotten private threatening emails telling me not
to attend public Apache events.
Not social. Check. Doesn't help the group find out about alternatives.
Check.
But it lowers the temperature on my side of the keyboard to not mention
MapR in public. I imagine that Tom feels much the same way.
Re: IBM big insights distribution
Posted by Michael Segel <mi...@hotmail.com>.
Why is it that when anyone asks a question about IBM Tom wants to take it off line?
Not very social.
I personally find it interesting that IBM requires that their products are tied to IBM's JVM.
I also wonder about the cost of running Hadoop on the mainframe.
(Linux LPARs while cheaper, still cost mips. ) [Forgive me, its been a while since I spoke mainframe]
So what's the real advantage from an architecture point of view?
On Sep 19, 2012, at 6:39 PM, Tom Deutsch <td...@us.ibm.com> wrote:
> Serge - let's take this off list. Ping me directly and we'll take it from there.
>
> ---------------------------------------
> Sent from my Blackberry so please excuse typing and spelling errors.
>
>
> ----- Original Message -----
> From: Serge Blazhiyevskyy [Serge.Blazhiyevskyy@nice.com]
> Sent: 09/19/2012 03:16 PM CST
> To: "user@hadoop.apache.org" <us...@hadoop.apache.org>
> Subject: IBM big insights distribution
>
>
>
> Hi all,
>
> Anybody had a success compiling Map reduce jobs with big insights distro on MAC?
>
> Seems like it requires IBM java JDK and it might not be available on MAC.
>
> Is there a way work around it?
>
>
> Thanks
> Serge
>
>
>
>
Re: IBM big insights distribution
Posted by Michael Segel <mi...@hotmail.com>.
Why is it that when anyone asks a question about IBM Tom wants to take it off line?
Not very social.
I personally find it interesting that IBM requires that their products are tied to IBM's JVM.
I also wonder about the cost of running Hadoop on the mainframe.
(Linux LPARs while cheaper, still cost mips. ) [Forgive me, its been a while since I spoke mainframe]
So what's the real advantage from an architecture point of view?
On Sep 19, 2012, at 6:39 PM, Tom Deutsch <td...@us.ibm.com> wrote:
> Serge - let's take this off list. Ping me directly and we'll take it from there.
>
> ---------------------------------------
> Sent from my Blackberry so please excuse typing and spelling errors.
>
>
> ----- Original Message -----
> From: Serge Blazhiyevskyy [Serge.Blazhiyevskyy@nice.com]
> Sent: 09/19/2012 03:16 PM CST
> To: "user@hadoop.apache.org" <us...@hadoop.apache.org>
> Subject: IBM big insights distribution
>
>
>
> Hi all,
>
> Anybody had a success compiling Map reduce jobs with big insights distro on MAC?
>
> Seems like it requires IBM java JDK and it might not be available on MAC.
>
> Is there a way work around it?
>
>
> Thanks
> Serge
>
>
>
>
Re: IBM big insights distribution
Posted by Michael Segel <mi...@hotmail.com>.
Why is it that when anyone asks a question about IBM Tom wants to take it off line?
Not very social.
I personally find it interesting that IBM requires that their products are tied to IBM's JVM.
I also wonder about the cost of running Hadoop on the mainframe.
(Linux LPARs while cheaper, still cost mips. ) [Forgive me, its been a while since I spoke mainframe]
So what's the real advantage from an architecture point of view?
On Sep 19, 2012, at 6:39 PM, Tom Deutsch <td...@us.ibm.com> wrote:
> Serge - let's take this off list. Ping me directly and we'll take it from there.
>
> ---------------------------------------
> Sent from my Blackberry so please excuse typing and spelling errors.
>
>
> ----- Original Message -----
> From: Serge Blazhiyevskyy [Serge.Blazhiyevskyy@nice.com]
> Sent: 09/19/2012 03:16 PM CST
> To: "user@hadoop.apache.org" <us...@hadoop.apache.org>
> Subject: IBM big insights distribution
>
>
>
> Hi all,
>
> Anybody had a success compiling Map reduce jobs with big insights distro on MAC?
>
> Seems like it requires IBM java JDK and it might not be available on MAC.
>
> Is there a way work around it?
>
>
> Thanks
> Serge
>
>
>
>
Re: IBM big insights distribution
Posted by Michael Segel <mi...@hotmail.com>.
Why is it that when anyone asks a question about IBM Tom wants to take it off line?
Not very social.
I personally find it interesting that IBM requires that their products are tied to IBM's JVM.
I also wonder about the cost of running Hadoop on the mainframe.
(Linux LPARs while cheaper, still cost mips. ) [Forgive me, its been a while since I spoke mainframe]
So what's the real advantage from an architecture point of view?
On Sep 19, 2012, at 6:39 PM, Tom Deutsch <td...@us.ibm.com> wrote:
> Serge - let's take this off list. Ping me directly and we'll take it from there.
>
> ---------------------------------------
> Sent from my Blackberry so please excuse typing and spelling errors.
>
>
> ----- Original Message -----
> From: Serge Blazhiyevskyy [Serge.Blazhiyevskyy@nice.com]
> Sent: 09/19/2012 03:16 PM CST
> To: "user@hadoop.apache.org" <us...@hadoop.apache.org>
> Subject: IBM big insights distribution
>
>
>
> Hi all,
>
> Anybody had a success compiling Map reduce jobs with big insights distro on MAC?
>
> Seems like it requires IBM java JDK and it might not be available on MAC.
>
> Is there a way work around it?
>
>
> Thanks
> Serge
>
>
>
>